Skip to main content

Full text of "Pharmacokinetics of plasmid DNA"

See other formats


PHARMACOKINETICS OF PLASMID DNA 



By 
BRETT EDWARD HOUK 



A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL 

OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT 

OF THE REQUIREMENTS FOR THE DEGREE OF 

DOCTOR OF PHILOSOPHY 

UNIVERSITY OF FLORIDA 

2000 



Copyright 2000 

by 

Brett Edward Houk 



This work is dedicated to my parents Nancy and Ronald Houk for all of their guidance 
throughout my life. 



ACKNOWLEDGMENTS 
I would like to acknowledge Dr. Jeffrey A. Hughes who, aside from my parents, 
has been the biggest influence in my life thus far. I would also like to acknowledge Dr. 
Guenther Hochhaus for his invaluable insight and guidance in this work. 



IV 



TABLE OF CONTENTS 

page 

ACKNOWLEDGMENTS iv 

LIST OF TABLES vii 

LIST OF FIGURES ix 

ABSTRACT xiv 

INTRODUCTION 1 

The Use of Naked pDNA as a Therapeutic Agent 2 

Effectiveness of Naked Plasmid DNA after Local Administration 3 

Effectiveness of Naked Plasmid DNA after IV Administration 6 

Anatomical Factors Potentially Involved in the Pharmacokinetics of Plasmid DNA .. 9 

Degradation of pDNA in the Bloodstream 16 

Pharmacokinetics of Liposomal DeUvery Vehicles 17 

Pharmacokinetics of Naked Plasmid DNA and Liposome: Plasmid DNA Complexes 

17 

Pharmacokinetics of Naked Plasmid DNA in the Blood after IV Injection 17 

Distribution of Plasmid DNA in Tissues after IV Injection 20 

Conclusions 21 

PHARMACOKINETICS OF PLASMID DNA IN ISOLATED RAT PLASMA 23 

Introduction 23 

Methods 25 

Theoretical 35 

Results 36 

Conclusions 46 

PHARMACOKINETICS OF PLASMID DNA AFTER IV BOLUS ADMINISTRATION 
IN THE RAT 52 

Introduction 52 

Methods 53 

Theoretical 57 

Results 60 

Conclusions 61 



DOSE DEPENDENCY OF PLASMID DNA PHARMACOKINETICS 71 

Introduction 71 

Methods 73 

Results 78 

Conclusions 90 

PHARMACOKINETIC MODELING OF PLASMID DNA AFTER IV BOLUS 
ADMINISTRATION IN THE RAT 102 

Introduction 102 

Theoretical 104 

Results 106 

Conclusions 114 

PHARMACOKINETICS OF LIPOSOME: PLASMID DNA COMPLEXES 122 

Introduction 122 

Methods 125 

Results 127 

Pharmacokinetics of Liposome:pDNA Complex Degradation in Rat Plasma ... 127 
Pharmacokinetics of Liposome:pDNA Complex Degradation in Rat Whole Blood 

128 

Pharmacokinetics of Liposome:pDNA Complexes after IV Bolus Administration 

in the Rat 129 

Conclusions 130 

CONCLUSIONS AND IMPLICATIONS 145 

Summary of Results 145 

Implications of Plasmid DNA Degradation in Isolated Plasma 145 

Comparison of /« Vitro and In F/vo Pharmacokinetics 146 

Effects of Increasing Dose of Plasmid DNA 148 

Results of the Curve Fitting Experiments 151 

Liposome: pDNA Complex Conclusions 152 

Future Directions 155 

Concluding Remarks 157 

LIST OF REFERENCES 159 

BIOGRAPHICAL SKETCH 166 



VI 



LIST OF TABLES 

Table page 

2-L Method parameters for pDNA analysis 34 

2-2. Pharmacokinetic parameters for pDNA in isolated rat plasma 41 

2-3. Noncompartmental parameters for pDNA after incubation in isolated plasma 42 

2-4. Pharmacokinetic parameters for pDNA after incubating the pGElSO plasmid in 

isolated rat plasma 49 

3-1. Pharmacokinetic parameters calculated after 500 |ag dose of SC pDNA 65 

3-2. Comparison of in vivo and in vitro pharmacokinetic parameters for pDNA 66 

3-3. Comparison of OC and L pDNA parameters after administration of SC pGL3, 

pGElSO, andpGeneMax 69 

4-1. Pharmacokinetic parameters estimated for supercoiled pDNA based upon the fit t=0 

concentration of SC pDNA 81 

4-2. Noncompartmental analysis of OC pDNA after IV bolus administration of SC 

pDNA 84 

4-3. Noncompartmental analysis of L pDNA after IV bolus administration of SC pDNA. .85 

4-4. Noncompartmental analysis of OC pDNA after IV bolus administration of OC 

pDNA at 2500 and 250 |j,g doses 89 

4-5. Noncompartmental analysis of L pDNA after IV bolus administration of L pDNA at 

2500 and 250 ^g doses 93 

5-1 . Pharmacokinetic parameters for pDNA based upon the model presented in the text. ..1 12 

5-2. Overall pharmacokinetic parameters for pDNA when all doses are fit 

simultaneously 113 



vn 



5-3. Pharmacokinetic parameters calculated after administration of OC pDNA at 2500 

and 250 |j,g doses 118 

5-4. Pharmacokinetic parameters calculated after administration of L pDNA at 2500 and 

250 ^g doses 119 

6-1. Noncompartmental analysis of pDNA after administration of liposome: SC pDNA 

complexes 137 

6-2. Comparison of SC pDNA pharmacokinetic parameters after administration of SC 
pDNA either in free form (naked) at 2500 |j,g dose or after administration as 
liposome: pDNA complexes at 500 |j.g dose 138 

6-3. Comparison of OC pDNA pharmacokinetic parameters after administration of SC 
pDNA either in free form (naked) or after administration as liposome: pDNA 
complexes at 500 |ig pDNA dose 139 

6-4. Comparison of L pDNA pharmacokinetic parameters after administration of SC 
pDNA either in free form (naked) or after administration as liposome: pDNA 
complexes at 500 ^g pDNA dose 140 



vni 



LIST OF FIGURES 

Figure page 

1-1. Potential sights for nicking of the phosphodiester backbone of DNA 10 

1-2. Model of plasmid DNA degradation in the bloodstream 11 

1-3. Schematic representation of pDNA (•) passing through a continuous capillary: (1) 
pinocytosis, (2) through intercellular junctions, and (3) passing through 
endothelial channels 12 

1-4. Schematic representation of pDNA (•) passing through a fenestrated capillary: (1) 
pinocytosis, (2) passing through a diaphragm fenestrae, and (3) passing through 
and open fenestrae 13 

1-5. Schematic representation of pDNA (•) passing through a discontinuous capillary. 

(1) pinocytosis and (2) passing through large pores in the endothelium 14 

2-1. Plasmid map of pGL3 Control 26 

2-2. Plasmid map of the pGeneMax-Luciferase 27 

2-3. Plasmid map of pGE 150 28 

2-4. Standard curve for supercoiled pDNA. Abcissa represents total ng estimated by UV 
absorbance at 260 nm. Ordinate represents ng calculated by the KDS ID 
software. Error bars represent ±1 standard deviation 31 

2-5. Standard curve for OC pDNA. Abcissa represents total ng estimated by UV 
absorbance at 260 nm. Ordinate represents ng calculated by the KDS ID 
software. Error bars represent ±1 standard deviation 32 

2-6. Standard curve for Linear pDNA. Abcissa represents total ng estimated by UV 
absorbance at 260 nm. Ordinate represents ng calculated by the KDS ID 
software. Error bars represent ±1 standard deviation 33 

2-7. Pharmacokinetic model of plasmid DNA degradation in rat plasma. The model is 
considered to be a unidirectional process. SC, OC, and L represent the amounts 
of supercoiled, open circular, and linear plasmid, respectively, in each 



IX 



compartment. The rate constants ks, ko, and ki represent the degradation constants 
for supercoiled, open circular, and linear plasmid, respectively 37 

2-8. Agarose gel analysis of pDNA degradation in isolated rat plasma. Lane 1; size 

standard, lane 2; 30 sec, lane 3; 1 min, lane 4; 2 min, lane 5; 3 min, lane 6; 5 min, 
lane 7; 10 min, lane 8; 20 min, lane 9; 30 min, lane 10; 45 min, lane 1 1; 60 min, 
lane 12; 80 min 39 

2-9. Experimental and fitted data based on the pharmacokinetic model described in the 
text. Data points represent actual experimental data. Lines represent values 
predicted by the model. Data represents mean of n=3 ± 1 standard deviation. 
Key: ^ supercoiled, • open circular, A linear 40 

2-10. Analysis of rate constants in dilute plasma. Degradation rate constants were 

modeled in PBS diluted rat plasma. Rate constants represent the fitted values of 
n=6 rats/ time point. Key: ♦ks in dilute plasma, Hko in dilute plasma. The value 
of ki is not reported due to the prolonged stability of linear plasmid in dilute 
plasma 44 

2-11. Degradation of supercoiled pDNA in 25% rat plasma after (A) incubating the 

plasma at 90°C for 10 min (B) the addition of 0.1 mM EDTA 45 

2-12. Comparison of concentrations of OC pDNA using (♦) pGE150 concentrations of 
OC pDNA using (■) pGL3 in isolated rat plasma. Data represents mean of n=3 
±1 standard deviation 47 

2-13. Comparison of concentrations of L pDNA using (♦) pGE150 versus 

concentrations of L pDNA using (■) pGL3 in isolated rat plasma. Data 

represents mean of n=3 ±1 standard deviation 48 

3-1. Photograph of the jugular cannula placement used for blood sampling 55 

3-2. Photograph of the femoral vein isolation and injection procedure used for IV bolus 

administration 56 

3-3. A representative gel from which plasmid amounts were quantified as described in 
the methods section. Lane 1 : size standard, lane 2: 1 min, lane 3: 2 min, lane 4: 
3.5 min, lane 5: 5 min, lane 6: 10 min, lane 7: 20 min, lane 8: 30 min, lane 9: 45 
min, lane 10: 60 min 59 

3-4. Experimental and fitted data based on the pharmacokinetic model described in the 
text. Data points represent actual experimental data. Lines represent values 
predicted by the model. Data represents mean of n=6 ± 1 standard devaition. 
Key: • open circular, ▲ linear 62 

3-5. Concentrations of OC pGL3 after ■: IV bolus administration of a 500 |ag dose of 

SC pGL3, and ♦: Incubation of SC pGL3 in isolated plasma at 37°C 63 



3-6. Concentrations of L pGL3 after ■: IV bolus administration of a 500 |ig dose of SC 

pGL3, and^: Incubation of SC pGL3 in isolated plasma at 37°C 64 

3-7. Concentrations of OC pDNA in the bloodstream after IV bolus administration of 
500 |j,g of pCMV-Luc, pGE150, or pGL3. Data represents mean of n=3 ± 1 
standard deviation 67 

3-8. Concentrations of L pDNA in the bloodstream after IV bolus administration of 500 
)ag of pCMV-Luc, pGE150, or pGL3. Data represents mean of n=3 ± 1 standard 
deviation 68 

4-1 Agarose gel analysis of pDNA after conversion to the OC form of the plasmid. Lane 
1 : Prior to treatment plasmid is predominately SC. Lane 2: After treatment 
plasmid is completely converted to to OC form 75 

4-2. Absorbance of pDNA before and after conversion to the OC form. Data represents 

averages of n=3 ± 1 standard deviation 76 

4-3. Agarose gel analysis of pDNA before and after conversion to the L form of the 
plasmid. Lane 1 : Size standard, Lane 2: before treatment the plasmid is 
predominately in the SC and OC form, Lane 3: Mixture of SC, OC, and L 
plasmid for reference. Lane 4: after treatment the plasmid is completely converted 
to the L form 77 

4-4. Concentrations of SC pDNA in the bloodstream after 2500 )ag dose. SC pDNA 

remained detectable through 1 minute after administration. Data points represent 
averages of n=3 ± 1 standard deviation. Lines represent a least squares fit of the 
data using the model described in the Methods section 80 

4-5. Concentrations of OC pDNA after IV bolus administration of: ■ 2500 i^g, ▲ 500 

l^g, # 333 |ig, or ♦ 250 |igof SCpDNA. Data represents mean of n=3 82 

4-6. Concentrations of L pDNA after IV bolus administration of: ■ 2500 |ag, A. 500 jag, 

• 333 ng, or ♦ 250 |ag of SC pDNA. Data represents mean of n=3 83 

4-7. Superposition of OC pDNA concentrations normalized for dose after administration 
of •: 2500 jag, ▲: 500 |ag, ♦: 333 [ig, or ■:250 |ag dose. Data represents mean 
of n=3 ± 1 standard deviation 86 

4-8. Concentrations of OC pDNA in the bloodstream after administration of OC pDNA 

at a 2500 |ag dose. Data represents mean of n=3 ± 1 standard deviation 87 

4-9. Concentrations of OC pDNA in the bloodstream after administration of OC pDNA 

at a 250 )j.g dose. Data represents mean of n=3 ± 1 standard deviation 88 

4-10. Concentrations of L pDNA in the bloodstream after administration of L pDNA at a 

2500 fj-g dose. Data represents averages of n=3 ± 1 standard deviation 91 



XI 



4-11. Concentrations of L pDNA in the bloodstream after administration of L pDNA at a 

250 i^g dose. Data represents averages of n=3 ± 1 standard deviation 92 

4-12. Area under the curve of OC pDNA after administration of a 2500 fj,g dose of SC or 
OC pDNA. Data represents mean of n=3 ± 1 standard deviation. * Indicates 
statistical significance by one way ANOVA (p<0.05) 94 

4-13. Area under the curve of OC pDNA after administration of a 250 |ig dose of SC or 
OC pDNA. Data represents mean of n=3 ± 1 standard deviation. * Indicates 
statistical significance by one way ANOVA (p<0.05) 95 

4-14. Area under the curve of L pDNA after administration of a 2500 |4,g dose of SC or 
OC pDNA. Data represents mean of n=3 ± 1 standard deviation. * Indicates 
statistical significance by one way ANOVA (p<0.05) 96 

4-15. Area under the curve of L pDNA after administration of a 250 )j,g dose of SC or 
OC pDNA. Data represents mean of n=3 ± 1 standard deviation. AUC 
differences were not statistically significant by one way ANOVA 97 

5-1. Model forpDNA clearance from the bloodstream 105 

5-2. Concentrations of (A) OC and (B) L pDNA in the bloodstream after 2500 fag dose 
of SC pDNA. Data points represent the averages of n=3 ±1 standard deviation. 
Lines represent concentrations predicted by the model 108 

5-3. Concentrations of (A) OC and (B) L pDNA in the bloodstream after 500 ]ig dose of 
SC pDNA. Data points represent the averages of n=3 ±1 standard deviation. 
Lines represent concentrations predicted by the model 109 

5-4. Concentrations of (A) OC and (B) L pDNA in the bloodstream after 333 |j.g dose of 
SC pDNA. Data points represent the averages of n=3 ±1 standard deviation. 
Lines represent concentrations predicted by the model 110 

5-5. Concentrations of (A) OC and (B) L pDNA in the bloodstream after 250 |j,g dose of 
SC pDNA. Data points represent the averages of n=3 ±1 standard deviation. 
Lines represent concentrations predicted by the model 11 1 

5-6. Concentrations of OC pDNA in the bloodstream after (A) 2500 |j.g and (B) 250 [ig 
dose of OC pDNA. Data points represent the averages of n=3 ±1 standard 
deviation. Lines represent concentrations predicted by the model 115 

5-7. Concentrations of L pDNA in the bloodstream after (A) 2500 |ag and (B) 250 \ig 
dose of OC pDNA. Data points represent the averages of n=3 ±1 standard 
deviation. Lines represent concentrations predicted by the model 116 



xu 



5-8. Concentrations of L pDNA in the bloodstream after (A) 2500 |ig and (B) 250 ng 
dose of L pDNA. Data points represent the averages of n=3 ± 1 standard 
deviation. Lines represent concentrations predicted by the model 117 

6-1. Liposome-pDNA complexes were incubated in rat plasma for various time points. 
10 1^1 of sample was loaded in each lane as described in the methods section. 
Lane 1; size standard, lane 2; 1 min, lane 3; 2 min, lane 4; 5 min, lane 5; 10 min, 
lane 6; 20 min, lane?; 30 min, lane 8; 60 min, lane9; 2 h, lane 10; 3 h, lane 1 1; 5.5 
h 131 

6-2. Agarose gel analysis of liposome/pDNA complexes. (A) 1:1 lipid:pDNA ratio, 

through 4 hours. (B) 3:1 lipid:pDNA ratio, through 6 hours. (C) 6:1 lipid:pDNA 
ratio, through 6 hours. *Indicates the 3 hour time point 132 

6-3. Lane 1: high molecular weight size standard, lane 2: 1:1 lipid:pDNA complexes, 

lane 3: 3:1 lipid:pDNA ratio (w/w), lane 4: 6:1 lipid:pDNA ratio 133 

6-4. (A)Degradation of SC pDNA in rat blood versus plasma. (B)Degradation of 

supercoiled pDNA in 3:1 and 6:1 (w/w) hposome/pDNA complexes incubated in 
heparinized rat whole blood. Error bars indicate standard deviation of n=3 rats 134 

6-5. Agarose gel analysis of pDNA after administration of liposome: pDNA complexes. 
Lane 1:15 sec, lane 2: 30 sec, lane 3: 45 sec, lane 4: 1 min, lane 5: 1.5 min, lane 
6: 2 min, lane 7: 2.5 min, lane 8: 3 min, lane 9: 4 min, lane 10: 5 min 135 

6-6. Plasma concentrations of SC, OC, and L pDNA after 500 \xg IV bolus 

administration of SC pDNA: liposome complexes. Key: ♦: SC, ■: OC, A: L 136 

7-1. Schematic representation of pDNA degradation in isolated plasma 147 

7-2. Schematic representation of pDNA pharmacokinetic parameters after IV bolus 

administration of SC pDNA in the rat 154 

7-3. Schematic representation of liposome pDNA clearance fi-om the bloodstream. In 
this model, removal fi-om the bloodstream of the lipid: pDNA complexes is 
assumed to be larger than the degradation of the complex 156 



xui 



Abstract of Dissertation Presented to the Graduate School 

of the University of Florida in Partial Fulfillment of the 

Requirements for the Degree of Doctor of Philosophy 

PHARMACOKINETICS OF PLASMID DNA 

By 

Brett E. Houk 

May 2000 

Chairman: Dr. Jeffrey A. Hughes 
Cochairman: Dr. Guenther Hochhaus 
Major Department: Pharmaceutics 

We sought to construct a complete pharmacokinetic model to describe the 
degradation of all three topoforms, supercoiled (SC), open circular (OC), and linear (L), 
of pDNA in vivo and in vitro. SC pDNA was incubated in isolated rat plasma at 37°C in 
vitro. At various time points, the plasma was assayed by electrophoresis for the amounts 
of SC, OC, and L pDNA remaining. The calculated amounts remaining were fit to linear 
differential equations describing this process. The calculated pharmacokinetic 
parameters suggested that SC pDNA degrades in isolated rat plasma with a half-life of 
1.2 min, OC pDNA degrades with a half-life of 21 min, and L pDNA degrades with a 
half-life of 1 1 min. Complexation of pDNA with cationic liposomes resulted in a portion 
of the supercoiled plasmid remaining detectable through 5.5 h in vitro. We next 
investigated the pharmacokinetics of SC plasmid DNA after IV bolus administration in 
the rat by following SC, OC, and L pDNA. SC pDNA was detectable in the bloodstream 
only after the highest, 2500 |j,g, dose and had a clearance of 390(±10) ml/min and volume 



xiv 



of distribution of 148(±26) ml. The pharmacokinetics of OC pDNA exhibited non-hnear 
characteristics with clearance ranging from 8.3(±0.8) to 1.3(±0.2) ml/min and a volume 
of distribution of 39(±19) ml. L pDNA exhibited linear kinetics and was cleared at 
7.6(±2.3) ml/min with a volume of distribution of 37(±17) ml. AUC analysis revealed 
60(±10) % of the SC was converted to the OC form, and nearly complete conversion of 
the OC pDNA to L pDNA. Clearance of SC pDNA was decreased after liposome 
complexation to 87(±30) ml/min. However, the clearance of OC and L pDNA was 
increased relative to naked pDNA at an equivalent dose to 37(±9) ml/min and 95(±37) 
ml/min, respectively. We conclude that SC pDNA is rapidly cleared from the circulation. 
OC pDNA displays non-linear pharmacokinetics. L pDNA exhibits first order kinetics. 
Liposome complexation protects the SC topoform, but the complexes are more rapidly 
cleared than the naked pDNA. 



XV 



CHAPTER 1 
INTRODUCTION 

Biotechnology is one of the most rapidly growing areas in the pharmaceutical 
sciences today. However, biotechnology products (e.g. proteins and peptides) suffer 
from poor stability, low absorption, and difficulties in delivery. It would therefore be 
ideal if the protein could be made in vivo, utilizing the body's own mechanisms to 
produce the competent protein. Gene therapy is one potential route by which to 
accomplish this goal. Gene therapy also offers the potential treatment of genetic 
diseases. The replacement of mutated, missing, or deleted DNA via gene therapy can 
result in the production of a competent protein. These potentials make gene therapy one 
of the most exciting and rapidly advancing areas of biotechnology. 

Early studies have revealed that systemically administered plasmid DNA (pDNA) 
can be expressed in animals (Kawabata et al. 1995; Mahato et al. 1995; Osaka et al. 1996; 
Song et al. 1997; Thierry et al. 1997) and humans (Valere 1999). Intravenous (IV) 
administration of DNA offers the potential advantage of allowing a wide distribution of 
activity in the body (Lew et al. 1995; Thierry 1995; Osaka et al. 1996; Thierry et al. 
1997). This route of administration allows the treatment of non-localized and systemic 
diseases. Previous research on the pharmacokinetics of non-viral gene therapies have 
only been observational citing that plasmid DNA degrades within 5 minutes after 
incubation in whole blood in vitro or after IV injection (Kawabata et al. 1995; Thierry et 
al. 1997). 



Plasmid DNA exists as three major topoforms. The native structure of non- 
damaged pDNA is supercoiled (SC). Single strand nicks to the phosphodiester backbone 
of pDNA yield an open circular (OC) form (Figure 1-1). This metabolite of SC pDNA is 
associated with significant transcriptional activity (-90-100%) (Adami et al. 1998; Niven 
et al. 1998). Further single strand nicks to the OC pDNA yield linear (L) pDNA, 
associated with a significant loss of activity (-90%). This process is schematically 
illustrated in Figure 1-2. 

In order to properly dose and achieve the desired levels of gene expression it will 
be necessary to understand the pharmacokinetics of pDNA. In initial human clinical 
trials with viral gene therapy, at least one study was terminated due to a patient death 
(Press 1999). This death was later attributed to the high doses utilized in the trails. Thus, 
the pharmacokinetics of pDNA is an essential area to be considered as gene therapy 
approaches clinical use. 

The Use of Naked pDNA as a Therapeutic Agent 

The use of naked pDNA as a drug after intravenous (IV) administration has been 
intensely investigated (Wang et al. 1995; Takeshita et al. 1996; Zhang et al. 1997; Budker 
et al. 1998; Song et al. 1998; Wang et al. 1998; Witzenbichler et al. 1998; Liu 1999; 
Zhang et al. 1999). The use of naked pDNA in vivo was initially reported after 
intramuscular (IM) or intradermal (SQ) administration in mammals (Wolff et al. 1991; 
Fazio et al. 1994; Katsumi et al. 1994; Bright et al. 1995; Donnelly et al. 1995; Lopez- 
Macias et al. 1995; Ulmer et al. 1995; Bright et al. 1996; Corr et al. 1996; Casares et al. 
1997; Danko et al. 1997; Lawson et al. 1997; Ragno et al. 1997; Haensler et al. 1999; 
Noll et al. 1999; Osorio et al. 1999). These studies have definitively shown efficient 
expression of a transgene can be achieved after administration of naked pDNA. The 



successes in these studies suggest that pharmacokinetic modehng of pDNA in the 
bloodstream after IV administration, or pDNA appearing in the bloodstream after local 
administration, is an area that must be more clearly defined in order to optimize gene 
therapy for clinical use. 
Effectiveness of Naked Plasmid DNA after Local Administration 

Fazio and coworkers (Fazio et al. 1994) demonstrated that a transgene could be 
efficiently secreted into the circulation after IM administration. Plasma accumulation of 
human Apo-E2 was demonstrated for at least 45 days after injection. After 
administration of pDNA encoding for an interferon transgene, interferons were detected 
fi-om days 7 to 28 post-DNA innoculation (Lawson et al. 1997). Administration of 
plasmid DNA encoding the chloramphenicol acetyltransferase gene (CAT) in sterile 
water lead to CAT transgene expression that peaked between 1 and 3 days and was 
detected up to 28 days after DNA administration. Together these results indicate that 
sustained expression can be obtained. 

Efficient immunization of monkeys, mice, dogs, and cats has been demonstrated 
using naked pDNA (Katsumi et al. 1994; Lopez-Macias et al. 1995; Ulmer et al. 1995; 
Bright et al. 1996; Ragno et al. 1997; Haensler et al. 1999; Noll et al. 1999; Osorio et al. 
1999). After injection of naked pDNA encoding for influenza hemagglutinin into the 
skin of mice and monkeys, induction of significant ELISA antibody titers and 
hemagglufination (HA) inhibition titers that were above the usual threshold values 
predictive of protection against influenza were demonstrated (Haensler et al. 1999). Mice 
immunized by various mucosal routes with a pDNA carrying the HA gene (pVlj- HA) 
induced a HA-specific cytotoxic T lymphocyte (CTL) response. Similarly, nasal 



immunization with the DNA vaccine induced primary CTLs against measles virus HA 
(Etchart et al. 1997). 

Plasmid DNA may also serve as an attractive means by which immunization to 
parasitic infection may be achieved. After injection of pDNA encoding for heat shock 
protein 65, T cell proliferation and antibodies to this protein were found to be elevated in 
rats when compared with both an arthritic control and naive animals (Ragno et al. 1997). 
A single immunization with pDNA encoding for Yersinia enterocolitica 60-kDa heat 
shock protein (Y- HSP60) was used for vaccination and induced significant Y-HSP60- 
specific T cell responses after 1 week (Noll et al. 1999). Induction of antibodies against 
Salmonella typhi OmpC porin by naked DNA immunization has also been demonstrated 
(Lopez-Macias et al. 1995). 

A pDNA expression vector encoding human factor IX as an example of 
immunogen was injected into mice three times at 10-day intervals (Katsumi et al. 1994). 
This resulted in production of antibodies to human factor IX. Spleen cells from 
inoculated mice also showed significant cytotoxic T lymphocyte response to target cells 
expressing human factor IX. Thus, IM and SQ injection of pDNA can induce immune 
responses against the encoded protein without an exposure to virus particles, and this 
approach may serve as the basis for immunotherapy in the treatment of cancer and 
infectious diseases in humans. 

Plasmid DNA encoding for viral proteins is also an attractive means by which 
immunization to viral infection may be achieved. The applicability of pDNA 
immunization technology for vaccine development was also investigated by immunizing 
dogs and cats by the IM and SQ routes with a pDNA vector encoding the rabies virus 



glycoprotein G (Osorio et al. 1999). The results demonstrated that non-facilitated, naked 
pDNA vaccines can eUcit strong, antigen-specific immune responses in dogs and cats, 
and DNA immunization may be a useful tool for future development of novel vaccines 
for these species. Plasmid DNA encoding for the large tumor antigen (T- Ag) of SV40 
was used to actively immunize mice to assess the induction of SV40 T-Ag-specific 
immunity (Bright et al. 1996). Direct injection of the recombinant SV40 T-Ag protein 
alone failed to induce SV40 T-Ag-specific CTL responses, whereas the pDNA encoding 
SV40 T-Ag elicited CTL activity specific for SV40 T-Ag. Naked pDNA induced 
immune responses that were protective against a lethal challenge with SV40-transformed 
cells. 

Naked pDNA has also been successful in the treatment of cancer by local 
administration. Direct intratumoral injection of free pDNA into mouse melanoma BL6 
solid tumor can also result in a high level of transfection. The average amount of 
chloramphenicol acetyltransferase (CAT) expressed by injecting 30 |a,g pDNA containing 
a CAT gene into a single BL6 tumor was 1 .9 +/- 1 .0 ng, which is comparable to that 
reported in the skeletal muscle (Yang and Huang 1996). An intratumoral injection of 
naked pDNA containing the HSV-TK gene (pAGO) resulted in tumor weight reduction 
(40-50%) in treated animals versus control groups. Moreover, histopathological analysis 
on tumors showed large areas of cavitary necrosis (85%) in treated groups compared to 
controls (10%)) (Soubrane et al. 1996). Thus direct injection of fi-ee pDNA may offer a 
simple and effective approach and might be a potential method for cancer gene therapy. 



Effectiveness of Naked Plasmid DNA after IV Administration 

Naked pDNA administration by IV injection has also been shown to be an 
effective means by which high levels of gene expression can be obtained (Wang et al. 
1995; Takeshita et al. 1996; Zhang et al. 1997; Budker et al. 1998; Song et al. 1998; 
Wang et al. 1998; Witzenbichler et al. 1998; Liu 1999; Zhang et al. 1999). Budker and 
coworkers demonstrated that pDNA can be delivered to and expressed within skeletal 
muscle of rats when injected rapidly, in a large volume (2 to 3 ml) (Budker et al. 1998). 

Liu and coworkers also showed naked pDNA can be efficiently expressed in mice 
(Liu 1999). As high as 45 |j,g of luciferase protein per gram of liver could be recovered 
by a single tail vein injection of 5 ^g of naked pDNA. Approximately 45% of 
hepatocytes expressed the transgene. Peak expression was obtained at 8 hours after 
administration and could be retained with repeated injections. 

Efficient naked pDNA expression has been obtained following delivery via the 
portal vein, hepatic vein, bile duct or direct IV administration via the tail vein in mice, 
rats, and dogs (Zhang et al. 1997; Zhang et al. 1999). The highest levels of expression 
were achieved after IV administration by rapidly injecting the pDNA in large volumes, 
approximately 2.5 ml. Over 15 [ig of luciferase protein/liver was produced in mice and 
over 50 )Lig in rats. Equally high levels of beta-galactosidase (beta-Gal) expression were 
obtained, in over 5% of the hepatocytes that had intense blue staining. Expression of 
luciferase or beta-Gal was evenly distributed in hepatocytes throughout the entire liver 
when either of the three routes were injected. Peri-acinar hepatocytes were preferentially 
transfected when the portal vein was injected in rats. These levels of foreign gene 
expression are among the highest levels obtained with nonviral vectors. Repetitive 



pDNA administration through the bile duct led to sustianed foreign gene expression. 
This study demonstrates that high levels of pDNA expression in hepatocytes can be 
easily obtained by IV injection. 

Takeshita and coworkers (Takeshita et al. 1996) investigated the hypothesis that 
naked pDNA encoding for vascular endothelial growth factor (VEGF) could be used in a 
strategy of arterial gene therapy to stimulate collateral artery development. Plasmid 
DNA encoding each of the three principle human VEGF isoforms (phVEGF121, 
phVEGF165, or phVEGF189) was applied to the hydrogel polymer coating of an 
angioplasty balloon and delivered percutaneously to one iliac artery of rabbits with 
operatively induced hindlimb ischemia. Compared with control animals transfected with 
LacZ, site-specific transfection of phVEGF resulted in augmented collateral vessel 
development documented by serial angiography, improvement in calf blood pressure 
ratio (ischemic to normal limb), resting and maximum blood flow, and capillary to 
myocyte ratio (suggesting increased vascularization). Similar results were obtained with 
phVEGF121, phVEGF165, and phVEGF189. This suggests that these isoforms are 
biologically equivalent with respect to in vivo angiogenesis. The potential for VEGF-C 
to promote angiogenesis in vivo was then tested in a rabbit ischemic hindlimb model 
(Witzenbichler et al. 1998). Ten days after ligation of the external iliac artery, VEGF-C 
was administered as naked pDNA (pcVEGF-C; 500 fig) from the polymer coating of an 
angioplasty balloon or as recombinant human protein (rhVEGF-C; 500 |j,g) by direct 
intra- arterial infusion. Physiological and anatomical assessments of angiogenesis 30 days 
later showed evidence of therapeutic angiogenesis for both pc VEGF-C and rh VEGF-C. 
Hindlimb blood pressure ratio (ischemic/normal) after pc VEGF-C increased after 



8 

pcVEGF-C versus controls and after rhVEGF-C versus control rabbits receiving rabbit 
serum albumin. Doppler- derived iliac flow reserve was increased for pcVEGF-C versus 
controls and increased for rhVEGF-C versus albumin controls. Neovascularity was 
documented by angiography in vivo after administration of pcVEGF-C and capillary 
density was measured at necropsy increased. Arterial gene transfer of naked pDNA 
encoding for a secreted angiogenic cytokine, thus, represents a potential alternative to 
recombinant protein administration for stimulating collateral vessel development. 

Naked pDNA constructs encoding for the human kallikrein protein delivered to 
spontaneously hypertensive rats via IV injection have been shown to be efficient at 
controlling hypertension (Wang et al. 1995). The expression of human tissue kallikrein in 
rats was identified in the heart, lung, and kidney by reverse transcription polymerase 
chain reaction followed by Southern blot analysis and an ELISA specific for human 
tissue kallikrein. A single injection of both human kallikrein pDNA constructs caused a 
sustained reduction of blood pressure, which began 1 week after injection and continued 
for 6 weeks. A maximal effect of blood pressure reduction of 46 mm Hg in rats was 
observed 2-3 weeks after injection with kallikrein pDNA as compared to rats with vector 
pDNA. These results show that direct gene delivery of human tissue kallikrein causes a 
sustained reduction in systolic blood pressure in genetically hypertensive rats and 
indicate that the feasibility of kallikrein gene therapy for treating human hypertension 
should be studied. 

Collectively, these results suggest that IV administration of naked pDNA is an 
attractive means to treat a large range of diseases. However, complete pharmacodynamic 
modeling of pDNA will has not been achieved. This will allow correlation of the 



administered dose with the desired levels of gene expression at the site of activity. 
Because of the plasmids high molecular weight, anatomical factors must be considered in 
the movement of these molecules within the body. 

Anatomical Factors Potentially Involved in the Pharmacokinetics of Plasmid DNA 

Plasmid DNA is a macromolecule having a molecular weight of 3.5 million for a 
typical plasmid of 5.5 kilobase pairs. This large molecular weight results in an increased 
likelihood of clearance processes being a function of size and its resulting abitily to pass 
through capillary endothelia. After IV administration, distribution of macromolecules is 
limited by the structure of the vascular endothehum. The structure of capillaries is 
diverse among organs. There are 3 main types of blood capillaries: continuous, 
fenestrated, and discontinuous (Hwang et al. 1997; Takakura et al. 1996). 

These 3 types of capillaries are represented in Figures 1-3, 1-4, and 1-5. The 
diameter of the free plasmid varies from between 8 to 22 nm (Yarmola 1985). The 
passage of pDNA through a continuous capillary would be limited to the 50 nm 
pinocytotic vesicles, 2 to 6 nm intracellular junctions, and 50 nm transendothelial 
channels (Figure 1-3) (Hwang et al. 1997). The basal lamina presents a barrier of 
collagen, glycoproteins, and fibronectin, macromolecules greater than 1 1 rmi can be 
retained by the basal lamina. Thus, this may present a barrier for difftision of the plasmid 
(Hwang et al. 1997). Continuous capillaries are the most widely distributed in 
mammalian tissue and are found in skeletal, cardiac, and smooth muscles, as well as lung, 
skin, subcutaneous tissues, serous membranes, and mucus membranes (Takakura et al. 
1996). 



10 



o 



^C^M 



'+° ^<x^^ 








M H 



O— p=o 



tr-' 



.1. 





H^ 



0-CH- 



h 



^. 










.-o Hc; 

-CHf 




GH H 




Figure 1-1. Potential sights for nicking of the phosphodiester backbone of DNA. 



11 



sc 



^ 



oc 






Endonuclease action: Endonuclease action: 
Single strand nick to Single strand nick 
the plasmid adjacent to previous 




Endonuclease or 
exonuclease action 



Figure 1-2. Model of plasmid DNA degradation in the bloodstream. 



12 



50 nm 




Figure 1-3. Schematic representation of pDNA (•) passing through a continuous 
capillary: (1) pinocytosis, (2) through intercellular junctions, and (3) passing through 
endothelial channels. 



13 



1 




Figure 1-4. Schematic representation of pDNA (•) passing through a fenestrated 
capillary: (1) pinocytosis, (2) passing through a diaphragm fenestrae, and (3) passing 
through and open fenestrae. 



14 




10^-10^ nm 





Figure 1-5. Schematic representation of pDNA (•) passing through a discontinuous 
capillary. (1) pinocytosis and (2) passing through large pores in the endothelium. 



15 



Fenestrated capillaries (Figure 1-4) are more likely to allow passage of pDNA 
into tissues. The pDNA may be transported through mechanisms similar to those 
involved in the continuous capillary, in addition to transport through 20 to 60 nm 
fenestrae (Hwang et al. 1997; Takakura et al. 1996). These fenestrae may or may not be 
closed by a diaphragm. The diameter of the closed diaphragm has not been reported. 
This type of capillary is generally found in the intestinal mucosa, endocrine glands, 
exocrine glands, glomerulus, and peritubular capillaries (Takakura et al. 1996). 

Discontinuous capillaries are characterized by endothelial gaps and large pores 
with diameters ranging from 100 to 1000 nm (Figure 1-5) (Hwang et al. 1997; Takakura 
et al. 1996). In these capillaries there is little restriction of diffusion of macromolecules. 
Another characteristic of this type of capillary is the lack of a basal lamina (Hwang et al. 
1997). The mucopolysaccharide rich interstitial Spaces of Disse have pore diameters 
ranging from 36 to 50 nm and are unlikely to present a major barrier for the transport of 
pDNA. The discontinuous capillary is more limited in its distribution than the other 
types and is found only in the liver, spleen, and bone marrow (Takakura et al. 1 996). 

These anatomical features can play an important role in the distribution of IV 
administered pDNA, and other macromolecules. In addition, capillary permeability can 
be ftirther enhanced in pathophysiological states such as cancer and inflammation 
(Takakura et al. 1996). Thus the fate of IV administered pDNA is determined not only 
by physio-chemical properties such as molecular weight, but also by anatomical features 
of the capillary endothelium present in each tissue. 



16 

Degradation of pDNA in the Bloodstream 

Early studies suggested that serum nucleases play a major role in the clearance of 
DNA from the bloodstream of injected animals (Gosse et al. 1965). Investigations by 
Chused and coworkers suggested that nucleases may not play a major role in the 
degradation of tritiated KB cell genomic DNA when IV injected in mice (Chused 1972). 
However, their assay was not able to identify the true activity of nucleases given that 
their assay utilized genomic DNA. Single strand cuts to the isolated genomic DNA 
would not yield small fragments and would be undetectable by their method. This would 
yield an underestimation of true nuclease activity. In contrast, single strand cuts to 
pDNA would lead to a degradation of the native SC structure to the OC form of the 
plasmid and be detectable by agarose gel analysis. 

Nucleases represent two subclasses of enzymes, endonucleases and exonucleases. 
Endonucleases act on the phosophodiester backbone of DNA in a continuous chain 
(Lodish 1995). Whereas, endonucleases act upon the free end (5' or 3') of the 
phosphodiester backbone in a linear segment of DNA. Investigations by Thierry and 
coworkers , utilizing agarose gel analysis, suggested that the main nuclease activity in the 
bloodstream was endonucleo lytic. This was based on the finding that the linear to 
supercoiled ratio increased with time and the SC: OC ratio remained identical to control 
(Thierry et al. 1997). However this view fails to recognize endonucleolytic activity on 
linear pDNA also generates degradation products. If endonuclease activity is the primary 
route of degradation, the kinetic ratios should all remain similar, owing to the fact that 
exonuclease activity would be masked by endonuclease activity. The pharmacokinetics 
of this degradation remain to be determined and may serve as a valuable tool in the 
understanding of the mechanisms of pDNA degradation observed in the bloodstream. 



17 

Pharmacokinetics of Liposomal Delivery Vehicles 
Although few studies are available on the pharmacokinetics of liposome:pDNA 
complexes, liposomal pharmacokinetics alone have been studied extensively with several 
reviews published (Hwang et al. 1997; Juliano 1988; Takakura et al. 1996). Liposome 
pharmacokinetics have been shown to be dependent upon size (Sato 1986), dose 
(Bosworth 1982; Osaka et al. 1996), lipid composition (Gabizon 1988), and charge 
(Juliano 1988). In general, liposomes larger than 60 nm in diameter are unable to access 
tissues having continuous capillary endothelia, including skeletal, cardiac, and smooth 
muscle, lung, skin, subcutaneous tissue, and serous and mucous membranes, and are 
limited to uptake in tissues of the reticuloendothelial system (Hwang et al. 1997). 
Liposomes larger than 0.5 |j,m are confined to the vasculature in all tissues. 

Pharmacokinetics of Naked Plasmid DNA and Liposome: Plasmid DNA Complexes 
After systemic administration of pDNA alone or as liposome:pDNA complexes, 
DNA rapidly disappears from the bloodstream. The processes responsible involve 
degradation in the blood stream, interaction with plasma proteins, organ distribution, and 
uptake by the reticuloendothelial system. The transport of DNA and liposome:pDNA 
complexes into organs is roughly a unidirectional system, where distribution back into 
the central compartment can be assumed to be negligible (Mahato et al. 1997). 
Pharmacokinetics of Naked Plasmid DNA in the Blood after IV Injection 

Plasma levels of pDNA may be measured using radiolabeled DNA or agarose gel 
analysis (Kawabata et al. 1995; Lew et al. 1995; Mahato et al. 1995; Osaka et al. 1996; 
Thierry et al. 1997). Using agarose gel analysis, Thierry and coworkers found that SC 
plasmid DNA is not detectable in either murine plasma or cell fractions 1 minute after 
injection of naked plasmid DNA in mice (Thierry et al. 1997). OC and L forms have 



18 

been detected through 30 minutes post-injection by Southern blot analysis (Lew et al. 
1995). The half-life of intact (OC or L) plasmid DNA is less than 5 minutes. Degraded 
plasmid fragments remain detectable in the blood at 30 minutes post injection. By 60 
minutes even degraded plasmid is cleared. This elimination has been shown to be 
independent of the DNA sequence (Lew et al. 1995). 

When plasmid DNA was administered to mice in the form of liposome:pDNA 
complexes, SC DNA was detected in the blood between 1 and 60 minutes after injection 
(Thierry et al. 1997). OC DNA degrades with a half-life of approximately 10 to 20 
minutes. Uptake of pDNA in blood cells reaches a maximum as early as 1 minute after 
injection of liposome:pDNA complexes. 

A major problem associated with these studies is that the analysis of samples was 
done only qualitatively. No attempt was made to quantitate the amounts and types of 
plasmid present in the bloodstream at various times. This information is critical for an 
evaluation of the predictive value of pharmacokinetic parameters associated with gene 
delivery. 

Quantitative analysis of gene delivery has been done using IV injected 
radiolabeled pDNA, [^^P] or [^^P], in mice. In these studies, the half-life of naked pDNA 
is approximately 10 minutes (Kawabata et al. 1995; Osaka et al. 1996). The total plasma 
radioactivity displays a degradation pattern consistent with a two compartment body 
model (Mahato et al. 1995). Total body clearance of naked pDNA is estimated at 102 
ml/hr, and plasma AUG is estimated at 0.98 (% of dose*hr/ml). Urinary radioactivity 
increases with time, indicating the degradation products are excreted via the kidney. 
Similar results were obtained following injection of radiolabeled liposome :pDNA 



19 

complexes with AUC's of 0.57 to 0.7 (% dose*hr/ml) and total clearance ranging from 
175.8 to 142.7 (ml/hr) (Mahato et al. 1995). Half-life for radiolabeled pDNA: 
dimethyldioctadecylammonium bromide: dioleoylphosphatidylethanolamine complexes 
was shorter ranging from 4 to 8 minutes (Osaka et al. 1996) suggesting rapid tissue 
entrapment of the liposome:pDNA complexes relative to naked plasmid. Twenty- four 
hours after injection, blood cell and plasma radioactivity for naked pDNA and 
liposome:pDNA complexes were similar (Osaka et al. 1996). 

Between these 3 analysis methods (agarose gel, Southern blotting, and 
radiolabeling) agarose gel analysis can determine more detailed information on the 
degradation of different structures of plasmid, (SC, OC, and L). This method is easy to 
apply and can be done under normal conditions without the limitations associated with 
radioactivity. The disadvantage of this method is that it is traditionally a semi- 
quantitative method. The advantage of the [^^P] and [^^P] methods is that these are 
quantitative methods and are more sensitive than agarose gel analysis. The disadvantages 
are that radiolabeling yields OC pDNA and thus, this method gives no information on the 
pharmacokinetics of SC pDNA. OC plasmid can also not be differentiated from L 
pDNA. The radioactivity is also counted without discriminating the degraded DNA 
fragments or the free label. Furthermore, special conditions and precautions are needed 
to handle radioactive materials. The difference between [ P] and [ P] is that [ P] has 
less personal danger and offers greater ease of handling than [ P] (Song et al. 1997; 
Nivenetal. 1998). 

Overall, when pDNA is injected in mice, SC pDNA has not been detected when 
administered as naked pDNA, but is after the injection of liposome:pDNA complexes. 



20 

After administration as naked pDNA, OC and L pDNA degrades with a half-life of 
between 5 and 10 minutes. The half-life of OC or L pDNA after administration of 
liposome:pDNA complexes ranges from 4 to 20 minutes. OC pDNA is available for 
transcription if taken up by cells. (Adami et al. 1998; Niven et al. 1998) Thus, 
administering pDNA in the form of liposome:pDNA complexes may offer a slight 
increase in the availability of IV administered pDNA. 
Distribution of Plasmid DNA in Tissues after IV Injection 

Tissue distribution of pDNA may be measured using radioactivity, Southern 
analysis, or whole body autoradiography. Using Southern analysis, pDNA has been 
detected in the bone marrow, heart, kidney, liver, lung, spleen, and muscle as early as 1 
hour after injection (Lew et al. 1995; Niven et al. 1998). No plasmid was detectable in 
the brain, intestine, and ovaries. 

Sub-picogram levels may be detected using polymerase chain reaction (PCR). 
Using this method, Lew and coworkers showed that at 7 days after IV injection, the range 
of residual plasmid was 1 fg/|ag in the brain, intestine, and gonads, and was 64 fg/|ig in 
the marrow, heart, liver, spleen, and muscle (translating to approximately 250-16,000 
copies/genome (Lew et al. 1995). By 28 days post-injection, levels of detectable plasmid 
had decreased 128 fold. Using PCR, residual plasmid remained detectable 6 months post 
injection at 2 to 8 fg/^ig genomic DNA and was predominantly in the muscle. 

After injection of radiolabeled plasmid, distribution may be measured by isolating 
tissues and measuring homogenates in a scintillation counter (Kawabata et al. 1995; 
Mahato et al. 1995). Alternatively, the entire carcass may be measured by whole body 
sectioning and autoradiography (Osaka et al. 1996; Niven et al. 1998). 



21 

After injection of radiolabeled naked pDNA, accumulation of radioactivity occurs 
initially in the lung, but declines rapidly through 1 minute post-injection (Kawabata et al. 
1995). Osaka and coworkers found that by 2 minutes after injection of naked pDNA, 
organ distribution is liver>spleen>lung, blood (Osaka et al. 1996). Whereas, Niven and 
coworkers found the time to reach maximum levels in the lungs is as long as 5 minutes 
versus 2 hours in the liver (Niven et al. 1998). Thus, there appears to be an initial rapid 
entrapment and transient accumulation in the lungs with accumulation occurring in the 
liver after a short period of time. Plasmid DNA was preferentially recovered in the non- 
parenchymal cells in the liver suggesting that the liver is acting in a scavenger role in 
uptake (Kawabata et al. 1995). 

When compared to naked pDNA, IV injection of liposome:pDNA complexes 
shows a higher accumulation of radioactivity in the lung 2 minutes after injection, Osaka 
and coworkers showed the major organs exhibit a distribution of 
lung>liver>spleen>kidney (Osaka et al. 1996). One hour after injection, a slight rise is 
seen in most organs, which is probably related to continuous uptake by the 
reticuloendothelial system. By 24 hours after injection of liposome:pDNA complexes, 
lung radioactivity dropped approximately 70 fold, with a distribution in major organs of 
spleen>liver>lung, kidney. 

Conclusions 

A complete understanding of the classical pharmacokinetic parameters of gene 
dehvery is necessary to move genetic agents forward as clinical therapeutics. Problems 
include the rapid clearance of naked pDNA and liposome:pDNA complexes without 
expression of the gene products, poor target tissue specificity, and degradation in the 
plasma. After systemic administration in mice, plasmid DNA is rapidly eliminated from 



22 

the circulation by extensive uptake by the reticuloendothelial system and degradation by 
plasma nucleases. Hepatic uptake is almost identical to liver blood flow suggesting 
highly efficient uptake. A complete pharmacokinetic model of all 3 forms of plasmid 
DNA (SC, OC, and L) has not been proposed. As the products of the biotechnology 
industry begin to move towards more clinical applications, the pharmacokinetic modeling 
of gene delivery will likely become an intensely investigated area. 



CHAPTER 2 
PHARMACOKINETICS OF PLASMID DNA IN ISOLATED RAT PLASMA 

Introduction 

In vivo delivery of plasmid DNA (pDNA) encoding for therapeutic proteins to 
patients via parenteral administration is an attractive means by which to target the gene to 
a wide variety of tissues. Early studies revealed that endogenous enzymes present in the 
plasma play a role in the clearance of nucleotides from the bloodstream (Chused 1972; 
Whaley 1972; Chia 1979; Piva 1998). These early studies have displayed that pDNA 
incubated in the presence of 10 % fetal bovine serum shows initial degradation by 15 min 
and is completely degraded by 60 min (Piva 1998). Similar results have been displayed 
in the presence of 90% human serum (Piva 1998). 

Nucleases will convert the native supercoiled (SC) pDNA topoform to the open 
circular (OC) and linear (L) forms of the plasmid (Lodish 1995). Changes in topoform 
have been associated with alterations in transcriptional activity. The significance of this 
change has been the matter of some debate. For example, the OC form of the plasmid 
has been shown to express similar levels of chloramphenicol acetyl transferase and 
luciferase proteins (Adami et al. 1998; Niven et al. 1998) to 2 to 4 times less (Hirose 
1993; Chemg 1999; Ramsey 1999) levels of transcribed luciferase and lac-Z, proteins. 
Increases in the amount of supercoiling serves to further increase the percent maximal 
transcription (Ramsey 1999). Furthermore, the time required for formation of the 
transcription preinitiation complex has been shown to be decreased with a SC template 
(Hirose 1993). Degradation to the L form of pDNA is associated with significant losses 

23 



24 

in transcriptional activity (90-100%) (Hirose 1993; Adami et al. 1998; Niven et al. 1998; 
Chemg 1999; Ramsey 1999). Differences in transcriptional activity may need to be 
accounted for in future pharamcodynamic studies. 

Early studies revealed that serum nucleases play a role in the rapid clearance of 
genomic DNA from the circulation of injected animals (Gosse et al. 1965). Recent 
studies on the pharmacokinetics of pDNA have attempted to use radiolabeled pDNA for 
detection (Osaka et al. 1996; Niven et al. 1998). However, the radiolabeling procedure 
involves nick translation, thereby eliminating the possibility of maintaining the SC 
topoform. Furthermore, this method does not discriminate the degraded pDNA from the 
intact plasmid, thus yielding an overestimation of the true half-life of the intact pDNA. 
Other studies on the pharmacokinetics of SC and OC pDNA have been only qualitative 
citing the presence of pDNA topoforms at various time points (Kawabata et al. 1995; 
Osaka et al. 1996; Thierry et al. 1997). 

Thierry and coworkers studied the stability of pDNA in the bloodstream of mice 
after IV injection (Thierry et al. 1997). Their results indicated that SC plasmid was not 
detectable in the plasma or red blood cell fractions 1 min after injection of pDNA. The 
true half-life was unable to be calculated using their method due to this rapid degradation. 
Kawabata and coworkers found that the SC pDNA was completely converted to the OC 
topoform within 5 min when incubated in mouse whole blood (Kawabata et al. 1995). 
Little other information on the pharmacokinetics of pDNA is available. The exact 
pharmacokinetics underlying this rapid degradative process is not fiilly understood. To 
properly dose and reach the desired therapeutic endpoints, a thorough understanding of 
the pharmacokinetics of pDNA is a necessity. 



25 

It is necessary to study the effects of plasma on pDNA in order to begin 
understanding the importance of the degradation of pDNA in the blood and allow a 
foundation upon which comparisons of delivery vehicles can be made. Naked pDNA has 
been shown to remain in the plasma fraction of blood (Osaka et al. 1996). For these 
reasons, we sought to investigate the pharmacokinetic processes underlying the stability 
of pDNA in a rat plasma model. We further sought to construct a complete 
pharmacokinetic model to describe the degradation of all three topoforms of pDNA in 
plasma. This model will allow a prediction of the time course of potential tissue 
exposure to the transcriptionally active SC and OC pDNA topoforms. 

Methods 

Phenol: chloroform: isoamyl alcohol (25: 24: 1, v/v/v), Tris, boric acid, EDTA, 
and agarose were purchased from Sigma Chemical Company (St. Louis, MO). Ethidium 
bromide (electrophoresis grade) was purchased from Fisher Biotech (Fair Lawn, NJ). 
Competent JM109 bacteria (Promega, Madison, WI) were transformed according to the 
manufacturers directions with the pGL3 control plasmid (Promega, Madison, WI), 
pGeneMax-Luciferase (Gene Therapy Systems, San Francisco, CA) or pGE150 plasmid 
(a generous gift of Dr. G. Elliot, Marie Curie Research Institute, The Chart, Oxted, 
Surrey, UK). Representative plasmid maps are presented in Figures 2-1, 2-2, and 2-3 for 
the pGL3, pGeneMax-Luciferase, and pGElSO plasmids, respectively. Plasmid DNA 
was isolated from overnight cultures using the Plasmid Maxi-Prep kit (Quiagen, 
Valencia, CA), and was >95% SC by agarose gel analysis. 

Blood was isolated from male Sprague-Dawley rats (300-350 g) by cardiac 
puncture, and immediately placed in heparinized test tubes (Vacutainer, Becton 



26 




Figure 2-1. Plasmid map of pGL3 Control. 



27 




Figure 2-2. Plasmid map of the pGeneMax-Luciferase. 



28 




Figure 2-3. Plasmid map ofpGElSO. 



29 

Dickinson, Franklin Lakes, NJ) on ice at the times indicated. Blood samples were 
centrifuged at 6,000 g for 5 min. For dilution experiments, plasma was diluted to 25 and 
50% with PBS or PBS containing 0.1 mM EDTA. To analyze the effects of heat, plasma 
samples were incubated at 90°C for 10 min in sealed tubes before assay. Plasma (600 |j.l) 
was removed and placed on ice until assay. Plasma samples were warmed to 37° C in a 
water bath and maintained at 37° C for the duration of the experiment. Plasmid DNA (12 
10.1/17 )j.g) in TE buffer was incubated in the 37° C plasma and 50 \x\ samples were taken 
at various times. Phenol: chloroform: isoamyl alcohol (25: 24: 1, v/v/v) (80 |a,l) was 
immediately added to each sample, vortexed for 5 s at low speed, and placed on ice. 
Samples were centrifuged at 20,800 x g for 10 min at room temperature. From the 
supernatant, an aliquot of 1 5 )j,1 was removed, 5 |a.l of 6 x loading dye (Promega, 
Madison, WI) added and placed on ice until loaded on an agarose gel. 

Samples were loaded on 0.8% agarose in 0.9 M Tris-Borate and 1 mM EDTA 
gels containing 0.001% ethidium bromide. Electrophoresis was carried out at 0.30 V/cm^ 
for 12 h. The ethidium bromide was excited on a UV light box (Model TFX-35M, Life 
Technologies, Grand Island, NY) and the net fluorescence intensity captured at 590 nm 
on a Kodak DC 120 digital camera (Eastman Kodak, Rochester, NY). The amounts of 
SC, OC, and L pDNA were calculated using Kodak Digital Science ID Image Analysis 
Software (version 3.0, Eastman Kodak Company, Rochester NY) with a Lambda Hind III 
digest (Promega, Madison WI) and a High DNA Mass Ladder (Life Technologies, Grand 
Island, NY) as external standards. Accuracy was 96 to 104% for SC pDNA, 98 to 108% 
for OC, and 96 to 1 13% for L pDNA. Percent coefficient of variation was < 5%, 19%, 
and 13% for SC, OC and L forms of the plasmid respectively. Standard curves for all 



30 

three forms of the plasmid (Figures 2-4, 2-5, and 2-6) were hnear between 10 and 250 ng 
pDNA bands (R^ = 0.9995, 0.9985, and 0.9933 for SC, OC, and L respectively). All 
reported concentrations were calculated from bands within the range of the standard 
curves. Lower limit of quantitation was 0.5 ng/|j,l. for all three forms of the plasmid. 
Lower limit of detection was 0.25 ng/|j,l for all three forms of the plasmid. Method 
parameters are summarized in Table 2-1. Comparisons of the relative fluorescence of SC 
pDNA versus OC and L pDNA were made by analyzing equivalent amounts as described 
above and comparing the resulting fluorescence. It was found that on a weight to weight 
ratio, SC pDNA was only 59% as fluorescent relative to L pDNA by agarose gel analysis. 
To correct for this difference, SC pDNA amounts were multiplied by 1.7 prior to 
analysis. This difference has been reported previously and is likely due to the relative 
inaccessibility of ethidium bromide to the SC topology (Cantor 1980). Percent recovery 
using the phenol: chloroform: isoamyl alcohol method was found not to be dependent on 
topoform. Recovery was 90 (± 6) % for SC and 86 (± 13) % for L pDNA. Comparisons 
of the relative fluorescence of SC pDNA versus L pDNA were made by digesting SC 
pDNA with the Hind III restriction enzyme (Promega, Madison, WI) which has a single 
recognition site in the plasmid. Equivalent amounts of L and SC pDNA were then loaded 
on agarose gels as described above and the relative fluorescence compared. Percent 
recovery was calculated by comparing phenol: chloroform: isoamyl alcohol (25: 24: 1, 
v/v/v) extracted versus non-extracted known amounts and analyzing on agarose gels as 
described above. 



31 



300 -, 



250 



200 



I 150 -i 



o 

8 



100- 



50 



R2 = 0.9995 



—I — 
50 



100 150 

A260 DNA equivalents 



200 



250 



Figure 2-4. Standard curve for supercoiled pDNA. Abcissa represents total ng estimated 
by UV absorbance at 260 nm. Ordinate represents ng calculated by the KDS ID 
software. Error bars represent ±1 standard deviation. 



32 



300 



250 - 



200 



a 

I 150-1 
o 

o 

100 A 



50 - 



R^ = 0.9985 



— I — 
50 



100 150 

A260 DNA Equivalents 



200 



250 



Figure 2-5. Standard curve for OC pDNA. Abcissa represents total ng estimated by UV 
absorbance at 260 run. Ordinate represents ng calculated by the KDS ID software. Error 
bars represent ±1 standard deviation. 



33 



300 ^ 



250 



200 



■o 

0) 
<-• 

■5 150 
o 

(0 

O 



100 - 



50 - 



R2 = 0.9933 



— r- 
50 



100 150 

A260 DNA equivalents 



200 



250 



Figure 2-6. Standard curve for Linear pDNA. Abcissa represents total ng estimated by 
UV absorbance at 260 nm. Ordinate represents ng calculated by the KDS ID software. 
Error bars represent ±1 standard deviation. 



34 



Table 2-1. Method parameters for pDNA analysis. 



Accuracy 


Supercoiled: 94-101 % 
Open Circular: 98-108% 
Linear: 96-113% 


Precision 


Supercoled: <5 % 
OpenCicular:<19% 
Linear: <13 % 


Lower Limit of Quantitation 


0.5 ng/|al 


Lower Limit of Detection 


0.25 ng/^1 


Recovery from Plasma 


Supercoiled: 90 (±6) % 
Linear: 86 (±13)% 



35 



Theoretical 

The degradation of SC pDNA was assumed to follow pseudo first-order kinetics. 
The model used is diagrammed in Figure 2-7. In this model, pDNA degradation is 
considered to be a unidirectional process. The degradation of L pDNA is considered to 
yield fragments of heterogeneous lengths, thus these products were not included in the 
fitted model. No elimination fi-om any of the compartments is assumed to occur through 
routes other than degradation to the following topoform. 

Based on this model the following differential equations were derived to describe 
the process: 



"'"^.-k.-SC 



dt 

dOC 

dt 



k-SC-k-OC 



— = k-OC-k,L 

dt " ' 



The amounts of supercoiled, open circular and linear pDNA were then fit to the 
integrated form of the equations: 



SC = SCq ■ e 



-k.t 



T -k -k • <>r ■( ' r"^°' I 1 .r''''' I ' ■r'''''\ 

^ % N ^'-O {{k,-k^Xk,-K) ^ ik„-k,Xk,-k,) ^ ^ (k„-k,)(k,-k,) ^ ) 

Where SC, OC, and L are the amounts of supercoiled, open circular, and linear pDNA 
present at time=t, respectively. SCo is the amount of supercoiled pDNA present at time 
(t)=0. The constants ko ,ks, and ki represent the rate constants for the degradation of SC, 
OC, and L pDNA respectively. The constants represent the activity of all enzymes acting 
in the degradation process. Non-linear curve fitting and goodness of fit, model selection 



36 

criteria (MSC) assessment was carried out using Scientist (version 4.0, Micromath, Salt 
Lake City, UT) (MicroMath 1995). Area under the plasma concentration time curve 
(AUC) was calculated using trapezoidal rule. Area under the terminal portion of the 
plasma concentration time curve, AUCtemi, was calculated by integration using the 
equation: 



AUC^ = 



Q 



last 



^term k 

Where Ciast is the last concentration point measured and k is the terminal elimination rate 

constant. Clearance (CI) was calculated from the volume (V) of rat plasma (7.8 ml) and 

the terminal elimination rate constant (k) using the equation (Davies 1993): 

Cl^V-k 

Statistical analysis was performed using SAS (The SAS Institute, Cary, NC). 

Results 
For quantitative purposes, the relative fluorescence of SC pDNA was compared to 
that of OC and L pDNA. It was found that on a weight to weight ratio, SC pDNA was 
only 59% as fluorescent relative to L pDNA by agarose gel analysis. To correct for this 
difference, SC pDNA amounts were multiplied by 1 .7 prior to analysis. This difference 
has been reported previously, and is likely due to the relative inaccessibility of ethidium 
bromide to the SC topology (Cantor 1980). Percent recovery using the phenol: 
chloroform: isoamyl alcohol method was found not to be dependent on topoform. 
Recovery was 90 (±6) % for SC and 86 (±13) % for L pDNA. 



37 



SC 



OC 



Linear 



:^> 



Figure 2-7. Pharmacokinetic model of plasmid DNA degradation in rat plasma. The 
model is considered to be a unidirectional process. SC, OC, and L represent the amounts 
of supercoiled, open circular, and linear plasmid, respectively, in each compartment. The 
rate constants ks, ko, and ki represent the degradation constants for supercoiled, open 
circular, and linear plasmid, respectively. 



38 



Figure 2-8 displays a representative gel in which the degradation of SC pDNA 
and the appearance of OC and L topoforms of plasmid is observed. In addition, the 
degradation products of L pDNA are visible as a light smear running below the band at 
60 min. Under the conditions used in this experiment, limit of quantification was 0.5 
ng/|j,l using the Lambda Hind III size standard. Plasmid amounts were calculated from 
agarose gel analysis using Kodak Digital Science ID image analysis software (Eastman 
Kodak, Rochester, NY) as described in the methods section. The observed and predicted 
values, based on the model displayed in Figure 2-8, are plotted in Figure 2-9. Plasmid 
concentrations were well described the model, MSC=3.0. Pharmacokinetic parameters 
calculated based on the model are summarized in Table 2-2. SC pDNA degraded rapidly 
in the plasma with a half-life of 1 .2 (± 0. 1 ) min. OC plasmid however was fairly stable, 
degrading with a half-life of 21 (± 1) min. L plasmid degraded more rapidly than the OC 
topoform but was fairly stable, in comparison to the SC plasmid degrading with a half- 
Ufe of 1 1 (± 2) min. OC AUC was nearly 17 times larger than SC, and 2.3 times larger 
than L pDNA (Table 2-3). 

No kinetics suggestive of enzyme saturation were observed under the 
experimental conditions tested. However, to ensure that saturation of plasma nucleases 
was not resulting in artificially low rate constant values, we analyzed the rate constants 
produced in dilute plasma (dilution was chosen as decreasing the dose of pDNA quickly 
results in a loss of sensitivity and sample sizes too large for loading). If saturation of 



39 



1 2 3 4 5 6 7 8 9 10 11 12 




Figure 2-8. Agarose gel analysis of pDNA degradation in isolated rat plasma. Lane 1 ; 
size standard, lane 2; 30 sec, lane 3; 1 min, lane 4; 2 min, lane 5; 3 min, lane 6; 5 min, 
lane 7; 10 min, lane 8; 20 min, lane 9; 30 min, lane 10; 45 min, lane 1 1; 60 min, lane 12; 
80 min. 



40 



"S> 



o 
a. 



15.0 



12.5 



10.0 



7.5 



5.0 



2.5 - 



0.0 




60 



time (min) 
Figure 2-9. Experimental and fitted data based on the pharmacokinetic model described 
in the text. Data points represent actual experimental data. Lines represent values 
predicted by the model. Data represents mean of n=3 ± 1 standard deviation. Key: ■ 
supercoiled, • open circular, ▲ linear. 



41 



Table 2-2. Pharmacokinetic parameters for pDNA in isolated rat plasma 



Topoform 


Rate 
constant 


Value (min' ) 


Standard 
Deviation 


Half-life (min) 


Supercoiled 


ks 


0.6 


0.03 


1.2 (±0.1) 


Open circular 


ko 


0.03 


0.002 


21 (± 1) 


Linear 


k, 


0.06 


0.008 


11 (±2) 


ita represent the 


itted values of n 


=6 ± 1 standard deviation. 





42 



Table 2-3. Noncompartmental parameters for pDNA after incubation in isolated plasma. 



Topoform 


AUC (ng/|il*inin) 


Clearance 
(fil/min) 


Supercoiled 


18 


360 (+ 9) 


Open circular 


310 


23 (± 1) 


Linear 


130 


47 (±5) 



Data represent n=6 ± standard deviation. AUC was calculated from the model fitted 
values using trapezoidal rule as described in the methods section. Clearance was 
calculated from the fitted rate constants and volume of rat plasma as described in the 
Methods section. 



43 



plasma nucleases was occurring, we expected that the rate constants in dilute plasma 
should deviate from a linear relationship negatively. Thus we tested the kinetics of 
pDNA degradation in 25% and 50% plasma. As displayed in Figure 2-10, no deviation 
was observed in the degradation of SC and OC pDNA. 

To further investigate the mechanism responsible for the observed degradation, 
and further validate our assay (to ensure the assay was not causing degradation itself) we 
studied the degradation of pDNA in PBS diluted 25% heated plasma (90°C for 10 min) 
and PBS containing 0.1 mM EDTA. No degradation of SC pDNA was observed in either 
case through 1 hr (Figure 2-11). The degradation sensitivity to heat and EDTA provides 
evidence that the degradation observed in the assay is due to enzymatic processes. 

We next sought to determine if the degradation observed in the previous 
experiments was dependent upon pDNA sequence. We, therefore, utilized the same 
model diagrammed above and replaced the pGL3 plasmid with the pGE150 plasmid. 
Unlike the pGL3 plasmid, which encodes for the luciferase protein and has an SV40 
promoter, this plasmid encodes for the green fluorescent protein and includes a CMV 
promoter. If the degradation of pDNA in the plasma was sequence dependent, we 
expected the degradation rate constants observed to differ from those observed in the 
previous experiment. A comparison of plasma concentrations of OC and L pDNA is 
presented in Figures 2-12 and 2-13. The resulting rate constants are presented in Table 4. 

Again the model diagrammed in Figure 2-7 described the data (model selection 
criteria = 3.1, rM.96, 0.99, and 0.92 for SC, OC, and L respectively). To determine if 



44 





0.7 . 
0.6 - 
0.5 - 

0.4 - 
0.3 - 
0.2 - 
0.1 - 

- 

( 




♦ 


^ 


♦ 


y = 0.6494x- 0.0311 
R2 = 0.9361^X^ 

y = 0.0415x- 0.0082 
R2 = 0.9962 




c 

1 


■ Ko 








) 0.2 




0.4 


% 


0.6 0.8 1 
plasma 


— 1 

1.2 



Figure 2-10. Analysis of rate constants in dilute plasma. Degradation rate constants were 
modeled in PBS diluted rat plasma. Rate constants represent the fitted values of n=6 rats/ 
time point. Key: ♦ks in dilute plasma, Bko in dilute plasma. The value of ki is not 
reported due to the prolonged stability of linear plasmid in dilute plasma. 



45 



B 




Std 0.5m Im 2m 3m 5m 10m 15m 20m 30m 45m Ihr Std 0.5m Im 2m 5m 10m 20m 30m 45m Ihr 



Figure 2-11. Degradation of supercoiled pDNA in 25% rat plasma after (A) incubating 
the plasma at 90°C for 10 min (B) the addition of 0.1 mM EDTA. 



46 

these rates were significantly different from those obtained using the pGL3 plasmid a 
statistical analysis was carried out using a 2-tailed equal variance student's t-test. The 
resulting parameters (ko, and ki) were not significantly different when judged at the 
p<0.05 criteria. These results suggest that pDNA sequence is not a major factor involved 
in the overall degradation of pDNA by plasma nucleases. 

Conclusions 
Previous reports on the pharmacokinetics of pDNA have only been qualitative, or 
involved radiolabeling. These studies indicated that pDNA degrades within 5 minutes in 
vitro or after IV injection (Kawabata et al. 1995; Thierry et al. 1997). In this study, we 
sought to quantitatively model the pharmacokinetics underlying the stability of pDNA in 
the plasma. The results revealed that SC pDNA degrades in the plasma with a half-life of 
1 min. OC pDNA is more stable than the SC topoform degrading with a half-life of 20 
min. L pDNA is degraded more rapidly than the OC topoform. This latter shortened 
stability is likely due to the accessibility of various nucleases present in the plasma to the 
L pDNA. OC plasmid must be nicked by endonucleases on each sister strand in the same 
location to generate L pDNA. However L pDNA would be accessible to both 
endonucleases and exonucleases, thus degrading more rapidly. The model and equations 
presented successfully described the degradation of pDNA in the plasma. 

Investigations by Thierry and coworkers suggested the main nuclease activity was 
endonucleolytic based on the finding that the L: SC ratio increased over time and the SC: 
OC ratio remained identical to control (Thierry et al. 1997). However this view fails to 



47 






Q 

Q. 








20 



40 



60 



80 



time (min) 



Figure 2-12. Comparison of concentrations of OC pDNA using (♦) pGE150 
concentrations of OC pDNA using (■) pGL3 in isolated rat plasma. Data represents 
mean of n=3 ±1 standard deviation. 



48 





10 


'^^ 




3 








O) 


1 


c 








< 

z 


0.1 


Q 




a 


rv r\A 



0.01 








20 



40 
time (min) 



60 



80 



Figure 2-13. Comparison of concentrations of L pDNA using (♦) pGE150 versus 
concentrations of L pDNA using (■) pGL3 in isolated rat plasma. Data represents mean 
of n=3 ±1 standard deviation. 



49 



Table 2-4. Pharmacokinetic parameters for pDNA after incubating the pGE150 plasmid 
in isolated rat plasma. 



Topoform 


Rate constant 


Value (min') 


Standard 
Deviation 


Half-life (min) 


Open Circular 


ko 


0.04 


0.007 


21 (±1) 


Linear 


k, 


0.06 


0.007 


11 (±1) 



Data represent the fitted values of n=6 rats. 



50 

recognize endonucleolytic activity on L pDNA also generates degradation products. If 
endouclease activity is the primary route of degradation, the kinetic ratios should all 
remain similar or decrease, owing to the fact that both exonucleases and endonucleases 
are active on the L pDNA and are thus both responsible for the observed degradation. 
Our results suggest that L pDNA has faster kinetics. This can be explained by 
endonuclease activity generating more free ends for degradation by exonucleases, 
exonucleases are more active than endonucleases, or that topoform influences the binding 
of these enzymes and thus influences reaction rate. Thus the main nuclease activity 
responsible for the observed kinetics remains to be answered. 

Area under the curve analysis revealed that tissues would be exposed to the OC 
topoform predominantly after injection of naked pDNA (Table 2-2). Blood flow through 
any individual organ becomes the limiting factor in its ability to uptake a drug, which is 
highly metabolized in the plasma. When compared to hepatic plasma flow in the rat 
(8.14 ml/min), clearance values for the degradation of SC plasmid (4.6 ml/min) suggest 
that metabolism in the bloodstream is a major pathway by which in vivo clearance of SC 
pDNA can occur (Davies 1993). However, given that the clearance by degradation in the 
plasma is less than the liver blood flow, it also suggests that the liver possess a perfusion 
rate sufficient for uptake of SC pDNA after IV injection. This parallels the findings of 
Kawabata and coworkers who observed that naked plasmid was cleared more rapidly 
from the circulation after IV injection than after in vitro incubation in whole blood 
(Kawabata et al. 1995). Lung, kidney, and spleen have also been shown to take up 
detectable amounts of plasmid after IV injection (Kawabata et al. 1995; Osaka et al. 



51 

1996). Our model establishes not only that tissue uptake of plasmid is possible, but also 
that tissue uptake of the non-nicked SC topoform is possible after IV injection. 
In summary this study presents a pharmacokinetic model describing the 
degradation of pDNA in rat plasma. A pharmacokinetic model is presented that can be of 
use in the future as gene therapy moves toward clinical trials. Using the derived model, 
we are able to conclude that naked SC pDNA degrades in rat plasma with a half-life of 
1.2 (± 0.05) min, OC with a half-Hfe of 21 (± 1) min, and L pDNA with a half-life of 1 1 
(± 2) min. 



CHAPTER 3 
PHARMACOKINETICS OF PLASMID DNA AFTER IV BOLUS ADMINISTRATION 

IN THE RAT 

Introduction 

Naked pDNA is being used successfully in gene delivery by administration EVI or 
SQ, (Haensler et al. 1999; Noll et al. 1999; Osorio et al. 1999; Rizzuto et al. 1999) and 
after IV injection (Wang et al. 1995; Budker et al. 1998; Song et al. 1998; Liu 1999; 
Zhang et al. 1999) in rats and mice. The success in these studies indicates that gene 
therapy is an attractive means by which to achieve therapeutic response. Thus, a 
thorough understanding of the pharmacokinetics of naked pDNA is an important area to 
be considered in order to move towards use in clinical trials 

The pharmacokinetics of pDNA after IV bolus administration have been 
investigated using radiolabeling with linearized ["P] pDNA (Osaka et al. 1996). These 
investigations have led to the conclusion that the half-life of the pDNA radiolabel is 7 to 
12 min after IV bolus administration of naked pDNA in mice. However, this analysis 
offers no information on the other functional forms of the plasmid; supercoiled (SC), 
open circular (OC), or linear (L), nor does it discriminate the fi-ee label. Other studies 
have qualitatively revealed that the SC topoform of pDNA is not detectable as early as 
one minute post IV injection in mice (Lew et al. 1995; Thierry et al. 1997). The OC form 
of the plasmid has a half-life estimated in these studies to be in the range of 10 to 20 
minutes (Thierry et al. 1997). 



52 



53 

The purpose of this investigation was to model the pharmacokinetics of naked 
pDNA in a topoform specific manner after IV bolus administration in the rat. We further 
sought to determine if the observed pharmacokinetics were affected by changes in 
plasmid sequence. These results were then compared to pDNA degradation in isolated 
plasma in order to determine the relative importance of plasma nucleases in the 
pharmacokinetics of pDNA. 

Methods 

Animals (male Sprague-Dawley rats 300-350 g) were purchased from Charles 
River Laboratories (Wilmington, MA). Animals were housed in the University of Florida 
Animal Resources Unit prior to all experiments and were given food and water ad 
libitum. Animals were anesthetized by intraperitoneal injection with 0.5 ml of a cocktail 
containing 13 mg/kg Xylazine, 2.15 mg/kg Acepromazine (The Butler Co., Columbus, 
OH), and 66 mg/kg Ketamine (Fort Dodge Animal Health, Fort Dodge Iowa). 

Phenol: chloroform: isoamyl alcohol (25: 24: 1, v/v/v), chloroform, Tris, boric 
acid, EDTA, and agarose were purchased from Sigma Chemical Company (St. Louis, 
MO). Ethidium bromide (electrophoresis grade) was purchased from Fisher Biotech 
(Fair Lawn, NJ). Competent JM109 bacteria (Promega, Madison, WI) were transformed 
according to the manufacturer's directions with the pGL3 control plasmid (Promega, 
Madison, WI) or pGeneMax plasmid (Gene Therapy Systems, San Francisco, CA). 
Plasmid DNA was isolated from overnight cultures using alkaline lysis and 
ultracentrifugation with a CsCl gradient (Katz 1977). Isolated pDNA was resuspended in 
phosphate buffered saline. Plasmid was >90% SC by agarose gel analysis. 

To facihtate blood sampling, male Sprague-Dawley rats (3 00-3 5 Og) were 
anesthetized and the jugular vein was exposed via an incision, isolated, ligated, and 



54 

nicked with ophthalmic scissors. A sterile silatstic (0.640 cm internal diameter by 0.12 
cm outer diameter, 10 cm in length) filled with sterile saline was threaded 30-40 mm into 
the jugular vein and positioned just distal to the entrance to the right atrium and secured 
by 6.0 silk sutures Figure 3-1). For injections, the femoral vein was isolated, and pDNA 
was injected into the femoral vein using a 27-gauge needle (Figure 3-2). Isolated blood 
samples (approx. 300 |il) were drawn through the jugular vein cannula and immediately 
placed in test tubes containing 0.57 ml of 0.34 M EDTA (Vacutainer, Becton Dickinson, 
Franklin Lakes, NJ) on ice at the times indicated. This concentration of EDTA has 
previously been shown to inhibit the degradation of pDNA in isolated rat plasma (Houk 
1999). 

To isolate pDNA fi-om whole blood samples, 250 |il of blood was liquid/ liquid 
extracted with 250 |il of phenol: chloroform: isoamyl alcohol (25: 24: 1, v/v/v), vortexed 
for 5 s at low speed, and centrifliged at 20,800 x g for 10 min at room temperature. The 
aqueous phase was removed and stored at -20°C until analysis. 

Samples were loaded on 0.8% agarose in 0.9 M Tris-borate, 1 mM EDTA gels 
containing 0.001% ethidium bromide. Electrophoresis was carried out at 0.30 V/cm^ for 
12h. The ethidium bromide was excited on a UV light box (Model TFX-35M, Life 
Technologies, Grand Island, NY) and the net fluorescence intensity captured at 590 nm 
on a Kodak DC 120 digital camera (Eastman Kodak, Rochester, NY). The amounts of 
SC, OC, and L pDNA were calculated using Kodak Digital Science ID Image Analysis 
Software (version 3.0, Eastman Kodak Company, Rochester NY) with a Lambda Hind III 
digest (Promega, Madison WI) and a High DNA Mass Ladder (Life Technologies, Grand 



55 




Figure 3-1 . Photograph of the jugular cannula placement used for blood sampling. 



56 




Figure 3-2. Photograph of the femoral vein isolation and injection procedure used for IV 
bolus administration. 



57 



Island, NY) as external standards. Accuracy was 96 to 104% for SC pDNA, 98 to 108% 
for OC, and 96 to 113% for L pDNA. Percent coefficient of variation was < 5%, 19%, 
and 13% for SC, OC and L forms of the plasmid respectively. Standard curves for all 
three forms of the plasmid (Figures 2-1, 2-2, and 2-3) were linear between 10 and 250 ng 
pDNA bands (R^ = 0.9995, 0.9985, and 0.9933 for SC, OC, and L, respectively). All 
reported concentrations were calculated from bands within the range of the standard 
curves. Lower limit of quantitation was 0.5 ng/|il for all three forms of the plasmid. 
Lower limit of detection was 0.25 ng/|al for all three forms of the plasmid. Method 
parameters are summarized in Table 2-1. Comparisons of the relative fluorescence of SC 
pDNA versus OC and L pDNA were made by analyzing equivalent amounts as described 
above and comparing the resulting fluorescence. It was found that on a weight to weight 
ratio, SC pDNA was only 59% as fluorescent relative to L pDNA by agarose gel analysis. 
To correct for this difference, SC pDNA amounts were multiplied by 1 .7 prior to 
analysis. This difference has been reported previously, and is likely due to the relative 
inaccessibility of ethidium bromide to the SC topology (Cantor 1980). Percent recovery 
using the phenol: chloroform: isoamyl alcohol method was found not to be dependent on 
topoform. Recovery was 90 (± 6) % for SC and 86 (± 13) % for L pDNA. 

Theoretical 
The degradation of SC pDNA was assumed to follow pseudo first-order kinetics. 
The model used is diagrammed in Figure 2-3. In this model, pDNA degradation is 
considered to be a unidirectional process. The degradation of L pDNA is considered to 
yield fragments of heterogeneous lengths, thus these products were not included in the 



58 

fitted model. No elimination from any of the compartments is assumed to occur through 
routes other than degradation to the following topoform. 

Based on this model the following differential equations were derived to describe 
the process: 



dSC 
dt 


-k^-SC 


dOC _ 

dt 


K-sc-k^- 


dt ' 


,OC-kiL 



The amounts of supercoiled, open circular and linear pDNA were then fit to the 
integrated form of the equations: 



k,t 



SC=SC,e-'' 

I -k -k SC ■( ' ■P'^°'4- ' .p~'''' + I r~''''] 

L. rt.^ /t^ OK.Q \(k,-k,)(k,-kj ^ ^ (k,-K)(k,-K) '^ ^ (k,~ki)(k,-k,) ^ ) 

Where SC, OC, and L are the amounts of supercoiled, open circular, and linear 
pDNA present at time=t, respectively. SCo is the amount of supercoiled pDNA present at 
time (t)=0. The constants ks ,ko, and k] represent the rate constants for the degradation of 
supercoiled, open circular, and linear pDNA, respectively. The constants represent the 
activity of all enzymes acting in the degradation process. Non-linear curve fitting and 
statistical analysis was carried out using Scientist (version 4.0, Micromath, Salt Lake 



59 



10 




Figure 3-3. A representative gel from which plasmid amounts were quantified as 
described in the methods section. Lane 1: size standard, lane 2: 1 min, lane 3: 2 min, lane 
4: 3.5 min, lane 5: 5 min, lane 6: 10 min, lane 7: 20 min, lane 8: 30 min, lane 9: 45 min, 
lane 10: 60 min. 



60 

City, UT). Noncompartmental pharmacokinetic analysis was carried out using standard 
parameters (Gibaldi 1982). 

Results 
SC pDNA was not detected as early as 30 seconds post-injection. The OC and L 

forms of the pDNA remained detectable through 30 minutes post-injection of the 500 )ig 
dose. An agarose gel analysis of the isolated samples is presented in Figure 3-3. 

An important parameter to be considered is the initial concentrations achieved 
after IV administration in comparison to the initial concentrations in vitro. The initial 
concentrations of SC pDNA in the in vitro experiments (i.e. the time=0 concentration) 
were 10 (± 0.3) ng/^il. After IV bolus administration, the initial extrapolated SC pDNA 
concentrations were 17 (± 5) ng/)j.l. Thus, we concluded that these concentrations were 
within a range relevant for comparison. The observed and fitted concentrations of OC 
and L pDNA are presented in Figure 3-4. The model again adequately described the 
data, model selection criteria=4.42. Pharmacokinetic parameters calculated based upon 
the model are presented in Table 3-1. 

A comparison of the in vitro and in vivo concentrations of OC and L pDNA are 
presented in Figures 3-5 and 3-6, respectively. Calculated pharmacokinetic parameters 
are presented in Table 3-2. OC pDNA half-life was markedly shorter after IV bolus 
administration than after incubation in isolated plasma, 5.3 (±1.4) versus 21 (±1) min. L 
pDNA removal was also more rapid after IV bolus administration, 1.9 (±0.8) versus 1 1 
(±2) min after incubation in isolated plasma. 

In order to ftirther investigate the importance of plasmid sequence on the observed 
pharmacokinetics, we injected the pGeneMax-Luciferase, and pGE150 plasmids by IV 



61 

bolus administration at equivalent dose (500 |ig). Concentrations of OC and L pDNA in 
the bloodstream are presented in Figures 3-7 and 3-8 respectively. The fitted elimination 
rate constants for OC and L pDNA were compared by 2-way ANOVA. The results are 
displayed in Table 3-3. There were no significant differences between the terminal rate 
constants of any of the 3 plasmids by 2-way ANOVA when judged at the p<0.05 criteria. 

Conclusions 
DNase I is a well characterized enzyme in human plasma present at 

concentrations averaging 26.1 (±9.2) ng/ml in the sera of normal humans (Chitrabamrung 
1981). Traditionally, the presence of this enzyme has led to the conclusion that pDNA 
administered IV is degraded rapidly (Gosse et al. 1965; Chused 1972). This has led to 
the current view of gene delivery, where protection fi"om plasma nucleases is a major 
goal of delivery systems. The results of this study demonstrate that although the half-life 
of SC and OC pDNA is remarkably short, degradation alone was not enough to explain 
the rapid disappearance of pDNA from the circulation observed in vivo. After IV bolus 
the rate of degradation of SC pDNA was greater than 7 times faster than in isolated 
plasma (Houk 1999). 

Chused and coworkers (Chused 1972) also suggested that nuclease activity was 
not enough to explain the rapid clearance of KB cell DNA fi-om the circulation in mice. 
In this study, only 2 to 3 % of the radioactivity was hydro lyzed to trichloroacetic acid 
(TCA) soluble fragments in 30 min, which was several half-lives longer than in the 
circulation. Tsumita and Iwanga (Tsumita and Iwanga 1963) also found that less than 5 
% of the total radioactivity was found in the TCA soluble fraction after 4.5 hours in 
mouse serum. 



62 



CD 

c 



Q 

Q. 



20 



15 - 



10 



5 - 




20 30 

time (min) 



40 



50 



Figure 3-4. Experimental and fitted data based on the pharmacokinetic model described 
in the text. Data points represent actual experimental data. Lines represent values 
predicted by the model. Data represents mean of n=6 ± 1 standard devaition. Key: • 
open circular, ▲ linear. 



63 








20 40 

time (min) 



60 



80 



Figure 3-5. Concentrations of OC pGL3 after ■: IV bolus administration of a 500 [ig 
dose of SC pGL3, and ♦: Incubation of SC pGL3 in isolated plasma at 37°C. 



64 



c 



Q 

Q. 



5 

4 
3 
2 
1 









20 40 

time (min) 



60 



80 



Figure 3-6. Concentrations of L pGL3 after ■: IV bolus administration of a 500 \xg dose 
of SC pGL3, and ♦: Incubation of SC pGL3 in isolated plasma at 37°C. 



65 



Table 3-1. Pharmacokinetic parameters calculated after 500 )j,g dose of SC pDNA. 



Topoform 


Rate 
constant 


Value 
(min') 


Standard 
Deviation 


Half-life 
(min) 


Supercoiled 


ks 


3.4 


0.4 


0.2 (± 0.03) 


Open circular 


ko 


0.14 


0.04 


5.3 (±1.4) 


Linear 


k, 


0.41 


0.18 


1.9 (±0.8) 



Parameters represent averages of n=6 rats. 



66 



Table 3-2. Comparison of /« vivo and in vitro pharmacokinetic parameters for pDNA. 



Topofonii 


Terminal Half-life 
(min) 

In vitro In vivo 


AUCoo (ng/^l*min) 
In vitro In vivo 


Cl/f(^l/mir 
In vitro 


) 
In vivo 


Supercoiled 


1.2 (±0.1) 


7 


17(±5) 


N/A 


360 (± 9) 


N/A 


Open Circular 


21 (± 1) 


5.3 (±1.4) 


280 
(±150) 


128 (± 52) 


23 (± 1) 


4800 
(±2000) 


Linear 


11(±2) 


1.9 (±0.8) 


103 (±47) 


49 (± 28) 


47 (± 5) 


11000 
(±5000) 



Parameters represent averages of n=3 (± 1 standard deviation) 



67 



100 





0.1 







5 10 

time (min) 



15 



Figure 3-7. Concentrations of OC pDNA in the bloodstream after IV bolus 
administration of 500 |ag of pCMV-Luc, pGE150, or pGL3. Data represents mean of n=3 
± 1 standard deviation. 



68 



O) 

c 



a 

Q. 



1 

0.1 


J' 


m-r- 




1^^^ 


0.01 




1 


: 1 








10 20 

time (min) 



30 



Figure 3-8. Concentrations of L pDNA in the bloodstream after IV bolus administration 
of 500 i^g of pCMV-Luc, pGElSO, or pGL3. Data represents mean of n=3 ± 1 standard 
deviation. 



69 



Table 3-3. Comparison of OC and L pDNA parameters after administration of SC pGL3, 
pGE150, and pGeneMax. 



Plasmid 


Parameter 


OC Value 


L Value 


pGL3 


AUCco 

(ng/^l*min) 


120 (±50) 


52 (±25) 


Cmax (ng/|ll) 


13 (±4) 


3.2 (±1.0) 


pGElSO 


AUCoo 
(ng/|il*min) 


160 (±30) 


55 (±12) 


Cmax (ng/|al) 


14 (±3) 


3.3 (±0.9) 


pGeneMax 


AUCoo 
(ng/^l*min) 


121 (±25) 


59 (±3) 


Cmax (ng/|J.l) 


12 (±5) 


3.7 (±0.3) 



Parameters represent averages of n=3 (± 1 standard deviation). 



70 

Alternatively, Gosse and coworkers suggested a major role for nucleases in the 
initial degradation of DNA after IV administration in rabbits and mice (Gosse et al. 
1965). This finding was based upon the proportionality between the initial rate of 
depolymerization and the plasma DNase activity level. Also, a rapid decrease in 
viscosity of isolated blood was discovered indicating a depolymerization of DNA. 
Finally, a markedly slower disappearance of DNA-methyl green complex (a non-specific 
DNase inhibitor) than after native DNA. 

The reason for this disparity in results deserves further investigation. Gosse 
ufilized much higher doses of pDNA than Chused and coworkers in their invesfigations, 
200 i^g versus 5 jag pDNA in Chused and coworkers 's investigations. This disparity 
may be due to saturation of a scavenger receptor, allowing nuclease activity to become 
increasingly important. The effect of increasing dose on the clearance of DNA deserves 
fiirther investigation. 

The results presented in the present study indicate that SC pDNA was 
undetectable after IV bolus administration, whereas SC pDNA was readily detectable in 
isolated plasma, and remained detectable through 3 min of incubation. Similar results 
were seen for the OC and L forms of the plasmid. The half-lives of OC and L pDNA 
decreased from 21 (±1) to 5.3 (±1.4) and 1 1 (±2) to 1.9 (±0.8) min, respecfively. Thus 
indicating that nuclease acfivity alone is not sufficient to describe the rapid clearance of 
pDNA from the bloodstream in rats. The observed kinetics were found not to be 
dependent upon plasmid sequence. 



CHAPTER 4 
DOSE DEPENDENCY OF PLASMID DNA PHARMACOKINETICS 

Introduction 

The studies presented in Chapters 2 and 3 have shown that degradation in the 
plasma alone was not sufficient to describe the pharmacokinetics of pDNA. After IV 
bolus administration SC pDNA was undetectable as early as 30 sec. This was in contrast 
to isolated plasma when SC pDNA was detectable 3 minutes after the start of incubation 
in isolated plasma. OC and L pDNA terminal half-life also decreased fi-om 21 (±1) to 5.3 
(±1.4) and 1 1 (±2) to 1.9 (±0.8) min, respectively. 

Other investigators have suggested variable importance of plasma nucleases in the 
degradation of genomic DNA after IV bolus administration. For example, Chused and 
coworkers (Chused 1972), Whaley and Webb (Whaley 1972), and Tsumita and Iwanaga 
(Tsumita 1963) all suggested a minimal role for plasma nucleases in the clearance of 
DNA. This was based upon the observed fragmentation rate of genomic DNA in isolated 
plasma versus the fragmentation rate after IV bolus administration, and the difftise high 
level of distribution of the DNA to tissues immediately after administration. This finding 
was accompanied by the suggestion of extensive uptake of intact DNA molecules by the 
reticuloendothelial system. 

Alternatively, Gosse and coworkers (Gosse et al. 1965) found that "the plasma 
DNases play a fiindamental and probably exclusive role in the initial degradation of 
DNA". This was based upon 3 observations. First was a rapid decrease in viscosity of 



71 



72 

the blood within 3 minutes after administration. Second, this was based upon the 
proportionahty between the initial rate of degradation and the DNase activity level. 
Third, this was also based upon the markedly slower disappearance of the DNA-methyl 
green complex (a non-specific DNase inhibitor). 

The disparity between the results presented here and the previous studies deserves 
fiirther investigation. Chused and coworkers, and Whaley and Webb, utilized smaller 
doses of DNA in their experiments versus Gosse and coworkers (5 versus 200 )j,g/ 
mouse). If this large dose had temporarily saturated an alternative clearance mechanism, 
this would increase the observed importance of nucleases. Thus, nonlinear processes 
may provide an explanation for the observed disparity. 

Nonlinear clearance of pDNA has previously been suggested using 
pharmacokinetic analysis of outflow patterns from rat perfiised liver studies with 
radiolabeled OC pDNA(Yoshida 1996). In this study, Vd increased and extraction ratio 
decreased as perfusion dose was increased from 1.33 to 13.3 i^g/liver. 

The purpose of this investigation was to model the pharmacokinetics of increasing 
doses of naked pDNA in a topoform specific manner after IV bolus adminisfration in the 
rat. This information may provide an explanation for the disparity between the results 
presented here and in previous studies. Furthermore, we sought to determine the 
metabolite (OC and L) pharmacokinetics independently, by direct injection of each of the 
metabolites. This informafion will provide a basis upon the percent conversion of the SC 
to the OC form and the OC form to the L form of the plasmid. 



73 

Methods 

Animals (male Sprague-Dawley rats 300-350 g) were purchased from Charles 
River Laboratories (Wilmington, MA). Animals were housed in the University of Florida 
Animal Resources Unit prior to all experiments and were given food and water ad 
libitum. Animals were anesthetized by intraperitoneal injection with 0.5 ml of a cocktail 
containing 13 mg/kg Xylazine, 2.15 mg/kg Acepromazine (The Butler Co., Columbus, 
OH), and 66 mg/kg Ketamine (Fort Dodge Animal Health, Fort Dodge Iowa). 

Phenol: chloroform: isoamyl alcohol (25: 24: 1, v/v/v), chloroform, Tris, boric 
acid, EDTA, and agarose were purchased from Sigma Chemical Company (St. Louis, 
MO). Ethidium bromide (electrophoresis grade) was purchased from Fisher Biotech 
(Fair Lawn, NJ). Competent JM109 bacteria (Promega, Madison, WI) were transformed 
according to the manufacturers directions with the pGL3 control plasmid (Promega, 
Madison, WI) or pGeneMax plasmid (Gene Therapy Systems, San Francisco, CA). 
Plasmid DNA was isolated from overnight cultures using alkaline lysis and 
ulfracentrifiigation with a CsCl gradient (Katz 1977). Isolated pDNA was resuspended in 
phosphate buffered saline. Plasmid was >90% SC by agarose gel analysis. 

OC pDNA was produced by incubation of the SC pDNA, in phosphate buffered 
saline, at 70°C for 16h. This procedure resulted in >90% OC plasmid (Figure 4-1). UV 
absorbance at 260 nm and the A260/A280 ratio of the pDNA solution did not change 
after this treatment (Figure 4-2). 

L pDNA was produced by digestion with BamHI restriction enzyme (Promega, 
Madison, WI) in separate reaction mixtures containing 173 [i\ of DI H2O, 27 ^1 lOx 
Buffer (Promega, Madison, WI) 56 ^1 (100 f^g) pGL3, and 10 ^1 of BamHI (10 U/j^l). 



74 

The reaction mixture was incubated at 37°C for 3 h. Plasmid was then isolated from the 
reaction mixture by extraction with 1 volume of phenol: chloroform: isoamyl alcohol (25: 
24: 1), followed by extraction with 1 volume of chloroform. Plasmid was then 
concentrated by precipitation with 0.3 M Na Acetate, and 1 volume of isopropanol, 
followed by centrifUgation at 13K g for 30 min at 4°C, and resuspended in 50 |j.l of 
phosphate buffered saline. This method routinely produced >90% L pDNA (Figure 4-3). 
Concentration of pDNA was measured by monitoring UV absorbance at 260 nm, purity 
was measured by A260/A280 ratio. A resulting purity of less than 1 .7 was re-extracted 
with 1 volume of chloroform until purity >1.7 was achieved. 

For blood sampling, male Sprague-Dawley rats (300-3 50g) were anesthetized and 
the jugular vein was exposed via an incision, isolated, ligated, and nicked with 
ophthalmic scissors. A sterile silatstic (0.64 cm internal diameter by 0.12 cm outer 
diameter, 10 cm in length) filled with sterile saline was threaded 30-40 mm into the 
jugular vein and positioned just distal to the entrance to the right atrium and secured by 
6.0 silk sutures. For injections, the femoral vein was isolated, and pDNA was injected 
into the femoral vein using a 27-gauge needle. This method is graphically illustrated in 
Figure 3-1 and 3-2. Isolated blood samples (approx. 300 ^1) were drawn through the 
jugular vein cannula and immediately placed in test tubes containing 0.57 ml of 0.34 M 
EDTA (Vacutainer, Becton Dickinson, Franklin Lakes, NJ) on ice at the times indicated. 
This concentration of EDTA has previously been shown to inhibit the degradation of 
pDNA in isolated rat plasma (Houk 1999). 



75 




<-OC 



<-sc 



Figure 4-1 Agarose gel analysis of pDNA after conversion to the OC form of the plasmid. 
Lane 1 : Prior to treatment plasmid is predominately SC. Lane 2: After treatment plasmid 
is completely converted to to OC form.. 



76 



I" 3500 1 



u 3000 

c 

8 

< 

z 
o 

a. 

■a 

I 2000 

3 
O 

« 



2500- 



1500- 

I 

"S 1000 

E 

1 500 
S 

« 




Supercoiled 



Open Circular 



Figure 4-2. Absorbance of pDNA before and after conversion to the OC form. Data 
represents averages of n=3 ± 1 standard deviation. 



77 







4-SC 



Figure 4-3. Agarose gel analysis of pDNA before and after conversion to the L form of 
the plasmid. Lane 1 : Size standard, Lane 2: before treatment the plasmid is 
predominately in the SC and OC form, Lane 3: Mixture of SC, OC, and L plasmid for 
reference. Lane 4: after treatment the plasmid is completely converted to the L form. 



78 

To isolate pDNA from whole blood 250 |il of blood was liquid/ liquid extracted 
with 250 f4.1 of phenol: chloroform: isoamyl alcohol (25: 24: 1, v/v/v), vortexed for 5 s at 
low speed, and centrifuged at 20,800 g for 10 min at room temperature. The aqueous 
phase was removed and stored at -20*'C until analysis. Samples analyzed and quantitated 
as described in Chapter 2. 

Results 

SC pDNA was detectable in the bloodstream only after a 2500 fj.g dose, no SC 
pDNA was detectable in the bloodstream at lower doses as early as 30 sec after 
administration. Because of this, the pharmacokinetic parameters reported for this form of 
pDNA relied only on data acquired from this dose. SC pDNA remained detectable in the 
plasma through 1 min after administration. Using the limited available data we 
approximated that SC pDNA degraded with a half-life of 0.15 (±0.01) min. The 
degradation of SC pDNA was fit to a one-compartment body model with central 
elimination and uptake (Figure 4-4). We extrapolated a least squares fit of the data to an 
initial, t=0, concentration which was necessary as this area accounted for a major portion 
of the AUCoo. Clearance of SC pDNA was calculated to be 390 (±10) ml/min, and 
volume of distribution was 148 (±26) ml. (Table 4-1). 

Concentrations of OC and L pDNA in the bloodstream after IV bolus 
administration are displayed in Figure 4-5 and 4-6 respectively. Noncompartmental 
analysis of all four doses is displayed in Table 4-2 for OC and Table 4-3 for L pDNA. A 
decrease in terminal slope is observable with increasing dose for the OC form of the 
plasmid (Figure 4-5). Clearance of the OC form of the plasmid decreased with increasing 
dose (Table 4-2). Formation clearance values for the OC form of the plasmid after the 



79 

administration of SC pDNA ranged from 1.3 (± 0.2) to 8.3 (± 0.8) ml/min for the 2500, 
and 250 |a.g doses respectively. Formation clearance of the L form of the pDNA 
remained constant at an average of 6.7 (± 0.2) ml/min for all doses. The 250 fig dose L 
concentrations close to limits of quantitation, and thus required a large amount of 
extrapolation for AUC calculation. For this reason, the 250 |ig dose L analysis was 
excluded from the noncompartmental analysis. Corresponding plots of OC pDNA 
plasma concentrations, normalized for dose, were not superimposable (Figure 4-7) 
(Gibaldi 1982). 

To investigate the percent of SC plasmid that becomes OC as well as the percent 
OC plasmid that becomes L, we compared the AUC obtained after IV bolus 
administration of the OC and L forms of plasmid independently at 2500 and 250 fig 
doses. Plasma concentrations of OC pDNA obtained after administration of 2500 and 
250 |ig doses are displayed in Figure 4-8 and 4-9 respectively. Noncompartmental 
analysis of the OC form of the plasmid at each dose is displayed in Table 4-4. Clearance 
again decreased between the 250 and 2500 jxg doses 8.8 (±2.4) to 1.3 (±0.2) ml/min. 
Clearance also remained consistent with that observed after administration of SC pDNA 
at each dose, 8.8 (± 2.4) versus 8.3 (±0.8) ml/ min at the 250 ^g dose, and 1.3 (± 0.2) 
versus 1 .3 (± 0.2) at the 2500 ]xg dose. Volume of distribution of the OC form was 43 
(±15) ml. 

Concentrations of L pDNA after administration of 2500 and 250 fag doses of L 
pDNA are presented in Figure 4-10 and 4-1 1 respectively. Noncompartmental analysis is 



80 



12 



10 



C 
< 

Z 
D 

Q. 



8- 



2- 









1 1 1 1 


1 1 1 1 1 1 1 1 1 1 1 r 1 I 1 ■■ I ■ ■■ T 1 1 

. . ; ; ,' ^ 




'■ 


■ 

; '■ 
; : ■ ' 

_ ; _ . . . 1. . 






1 1 1 1 1 1 1 1 1 1 1 [ T -T ip 1 1 1 1 1 



0.0 



0.5 



1.0 



1.6 



2.0 



time (min) 

Figure 4-4. Concentrations of SC pDNA in the bloodstream after 2500 ng dose. SC 
pDNA remained detectable through 1 minute after administration. Data points represent 
averages of n=3 ± 1 standard deviation. Lines represent a least squares fit of the data 
using the model described in the Methods section. 



81 



Table 4-1. Pharmacokinetic parameters estimated for supercoiled pDNA based upon the 
fit t=0 concentration of SC pDNA 



Parameter 


Value 


AUC (ng/^l*min) 


6.4 (±0.2) 


MRT(min) 


0.21 (±0.02) 


CI (ml/min) 


390 (±10) 


Vdss (ml) 


148 (±26) 


Half-life (min) 


0.15 (±0.02) 



Parameters represent averages of n=3 ±1 standard deviation. 



82 



D) 

c 



Q 

a 



100 



10 



0.1 




10 20 30 40 

time (min) 



50 



60 



Figure 4-5. Concentrations of OC pDNA after IV bolus administration of: ■ 2500 \xg, A 
500 i^g, • 333 ^g, or ♦ 250 |ig of SC pDNA. Data represents mean of n=3. 



83 



10 








10 20 30 40 

time (min) 



50 



60 



Figure 4-6. Concentrations of L pDNA after IV bolus administration of: ■ 2500 fxg, ▲ 
500 |ig, • 333 (ag, or ♦ 250 fig of SC pDNA. Data represents mean of n=3. 



84 



Table 4-2. Noncompartmental analysis of OC pDNA after IV bolus administration of SC 
pDNA. 



Parameter 


2500 lag 
Dose 


500 ^g 
Dose 


333 ^g 
Dose 


250 Mg 
Dose 


AUC 
(ng/)j,l*min) 


1200 
(±200) 


120 (±50) 


59 (±3) 


18 (±2) 


AUC % extrapolated 


1 (±0.4) 


9 (±4) 


10 (±6) 


24 (±2) 


AUMC 
(ng/(il*min^) 


20000 
(±6000) 


1900 
(±1200) 


400 (±20) 


130 (±20) 


MRT (min) 


16 (±3) 


14 (±3) 


6.8 (±0.4) 


7.2 (±0.3) 


Cl/f(ml/min) 


2.1 (±0.4) 


4.8 (±2.0) 


5.7 (±0.3) 


14 (±1) 


CI (ml/min) 


1.3 (±0.2) 


3.0 (±1.2) 


3.5 (±0.2) 


8.3 (±0.8) 


Cmax (ng/^1) 


49 (± 4) 


13 (±4) 


6.5 (±0.3) 


2.2 (±0.2) 


tmax (min) 


1 


1 


0.7 (± 0.3) 


0.8 (± 0.3) 



Parameters represent averages of n=3 (± 1 standard deviation). 



85 



Table 4-3. Noncompartmental analysis of L pDNA after IV bolus administration of SC 
pDNA. 



Parameter 


2500 ^g 
Dose 


500 ^ig 
Dose 


333 ^g 
Dose 


AUC 
(ng/|il*min) 


240 (±40) 


52 (±25) 


32 (±5) 


AUC % extrapolated 


12 (±7) 


15 (±5) 


13 (±7) 


AUMC 
(ng/^l*min^) 


7500 
(±2700) 


570 
(±370) 


300 (±20) 


MRT (min) 


31 (±6) 


10 (±2) 


9.6 (±1.7) 


Cl/f(ml/min) 


10.6 (±2.0) 


11 (±5) 


11 (±1) 


CI (ml/min) 


6.5 (±1.2) 


6.9 (+7.8) 


6.6 (±0.9) 


Cmax(ng/|al) 


5.4 (± 0.6) 


3.2 (±1.0) 


2.4 (±0.5) 


tmax (min) 


22 (± 3) 


5.3 (± 4.0) 


6.0 (±3.6) 



Parameters represent averages of n=3 (± 1 standard deviation). 



86 



0.03 








20 40 

time (min) 



60 



80 



Figure 4-7. Superposition of OC pDNA concentrations normalized for dose after 
administration of •: 2500 ^ig, ▲: 500 |ig, ♦: 333 |j,g, or ■:250 fjg dose. Data represents 
mean of n=3 ± 1 standard deviation. 



87 





70 




60 


3 


50 


c 


40 


< 

z 

Q 
a 


30 
20 
10 









10 20 30 40 50 60 70 80 90 
time (min) 



Figure 4-8. Concentrations of OC pDNA in the bloodstream after administration of OC 
pDNA at a 2500 |ig dose. Data represents mean of n=3 ± 1 standard deviation. 



88 



C 



O 

a 



5 
4 
3 
2 
1 




1^^ 

1^^- 








^^^^ 



i-^ 



^ 



10 
time (min) 



15 



20 



Figure 4-9. Concentrations of OC pDNA in the bloodstream after administration of OC 
pDNA at a 250 |ag dose. Data represents mean of n=3 ± 1 standard deviation. 



89 



Table 4-4. Noncompartmental analysis of OC pDNA after IV bolus administration of OC 
pDNA at 2500 and 250 ^g doses. 



Parameter 


2500 ^g 


250 ^ig 


AUC (ng/^l*min) 


1900(1200) 


30 (±9) 


AUC % extrapolated 


<1 


15 (±2) 


AUMC (ng/^iPmin') 


50000 (±5000) 


220 (±140) 


MRT (min) 


22 (±1) 


6.8 (±2.3) 


CI (ml/min) 


1.3 (±0.2) 


8.8 (±2.4) 


Vd,, (ml) 


29 (±3) 


56 (±5) 



Parameters represent averages of n=3 ± 1 standard deviation 



90 

presented in Table 4-5. The data was consistent with the clearance values observed after 
administration of SC pDNA, 7.6 (± 2.4) versus 6.6 (± 1.6). Volume of distribution for 
the L form of the plasmid was 38 (±12) ml. 

As is displayed in Figures 4-12 and 4-13, OC AUC after administration of SC 
pDNA was only 64 (± 11)% and 59 (± 1 1) % of the AUC after administration of OC 
pDNA for the 2500 and 250 ^g doses respectively. The AUC of L pDNA after 
administration of OC pDNA was 105 (± 23) and 95% (± 1 1) of the AUC of the AUC 
after administration of L pDNA for the 2500 and 250 |ig doses respectively (Figures 4-14 
and 4-15 respectively). 

Conclusions 

The results of this study reveal that all forms of pDNA (SC, OC, and L) are 
rapidly cleared from the circulation. Other investigators have qualitatively commented 
on the rapid clearance observed after IV bolus administration of pDNA (Lew et al. 1995; 
Mahato et al. 1995; Thierry et al. 1997). However, these studies have been limited to the 
OC and L forms of the plasmid. The half-life of the SC topoform has been unable to be 
estimated due to lack of detection (Lew et al. 1995; Thierry et al. 1997). Thierry and 
coworkers (Thierry et al. 1997) utilized electrophoresis and estimated the half-life of the 
OC form of the plasmid to be in the range of 10 to 20 min at a dose of 3.5 ^g/g in mice. 
This corresponds to a dose of approximately 1 100 |ag in a rat and is in reasonable 
agreement with the terminal half-lives that we observed here between the 500 and 2500 
l^g doses. Osaka and coworkers (Osaka et al. 1996) utilized a dose of 2.25 |ag/g of 
linearized radiolabeled plasmid and found the half-life to be 6.6 and 1 1.5 min (n=2). This 
corresponds to an approximate dose of 730 [ig in a rat. We found the half-life of 



91 








5 10 

tim e (m in) 



15 



Figure 4-10. Concentrations of L pDNA in the bloodstream after administration of L 
pDNA at a 2500 |ag dose. Data represents averages of n=3 ± 1 standard deviation. 



92 






Q 

Q. 



3 
2 
1 




-i-i 



■t 



1 



2 4 

time (min) 



Figure 4-11. Concentrations of L pDNA in the bloodstream after administration of L 
pDNA at a 250 ^g dose. Data represents averages of n=3 ± 1 standard deviation 



93 



Table 4-5. Noncompartmental analysis of L pDNA after IV bolus administration of L 
pDNA at 2500 and 250 [ig doses 



Parameter 


2500 i^g 


250 ^ig 


AUC (ng/^l*min) 


330 (± 40) 


22 (±4) 


AUC % extrapolated 


3.0 (±1.0) 


51 (±16) 


AUMC (ng/^iPmin') 


1500 (±200) 


77 (±24) 


MRT (min) 


4.5 (±0.1) 


3.5 (±1.2) 


CI (ml/min) 


7.6 (±0.8) 


11 (±2) 


Vdss (ml) 


34 (±4) 


51 (±17) 



Parameters represent averages of n=3 ± 1 standard deviation. 



94 



2500 

£ 2000 

E 

3 1500 

c 1000 



o 

< 



500 






OC AUG after OC 



OC AUG after SC 



Figure 4-12. Area under the curve of OC pDNA after administration of a 2500 ^ig dose 
of SC or OC pDNA. Data represents mean of n=3 ± 1 standard deviation. * Indicates 
statistical significance by one way ANOVA (p<0.05). 



95 





50 


^.^ 




c 


40 


E 




3 


30 






c 


20 


■^^ 




O 

D 


10 


< 











■ m 



OC AUG after OC 



OC AUG after SG 



Figure 4-13. Area under the curve of OC pDNA after administration of a 250 ng dose of 
SC or OC pDNA. Data represents mean of n=3 ± 1 standard deviation. * Indicates 
statistical significance by one way ANOVA (p<0.05). 



96 



D) 

c 
O 
< 




L AUG after L 



L AUG after L AUG after SG 
OG 



Figure 4-14. Area under the curve of L pDNA after administration of a 2500 \ig dose of 
SC or OC pDNA. Data represents mean of n=3 ± 1 standard deviation. * Indicates 
statistical significance by one way ANOVA (p<0.05). 



97 





L AUG after L 



L AUG after OG 



Figure 4-15. Area under the curve of L pDNA after administration of a 250 )j,g dose of 
SC or OC pDNA. Data represents mean of n=3 ± 1 standard deviation. AUC differences 
were not statistically significant by one way ANOVA. 



98 

linear pDNA to be 1 .7 (± 0.4) in our experiments. This difference may be due to species 
variation or an inability to differentiate the free radiolabel. 

The SC form of the plasmid disappeared rapidly from the circulation. One 
possible explanation for the rapid disappearance of SC pDNA is that it is rapidly being 
converted to the OC form in vivo by endogenous plasma nucleases. Traditionally the 
presence of DNase I, which is present at concentrations averaging 26.1 (± 9.2) ng/ml in 
the sera of normal humans (Chitrabamrung 1981), has led to the conclusion that pDNA 
administered IV is degraded rapidly [Gosse, 1965 #20; Chused, 1972 #19]. If we 
compare the reported concentrations of DNase I in human plasma along with the reported 
SC pDNA nicking activity of DNase I under optimal conditions (Dwyer 1999), and make 
the assumption that the activity of rat DNase I is approximately the same as the human 
isoform (Takeshita et al. 1996), we can arrive at an approximate activity of 0.1 (ng 
pDNA/^I/min). This is far less than the in vivo SC pDNA nicking rate of 9.2 (ng 
pDNA/^l/min) observed 30 sec after administration of the 2500 [ig dose. One minute 
after administration the in vivo rate was 2.7 ng pDNA/)il/min. Thus, the activity of 
DNase I would not seem to be enough to describe the rapid conversion of SC pDNA to 
the OC form. The combined effect of enzymes in addition to DNase I is also insufficient 
to describe the kinetics in total. After IV bolus administration, the rate of degradation of 
SC pDNA was greater than 7 times faster than in isolated rat plasma. 

Furthermore, if SC were being rapidly metabolized to the OC form, the AUC of 
OC pDNA after administration of SC pDNA would be nearly equal to the AUC after 
administration of OC pDNA. However, the AUC of OC was only 60 (± 10) % of the 



99 

AUC obtained after OC, leaving -40% of the plasmid to be accounted for by entrapment, 
association, or conversion to a form undetectable using this method. 

Open circular pDNA was cleared less rapidly than the SC form of the plasmid. 
This may be partially due to the accessibility of various nucleases. Open circular plasmid 
must be nicked by endonucleases on each sister strand in the same location to generate L 
pDNA. It is reasonable to assume that there are single strand nicks occurring in the OC 
plasmid during its entire time course of elimination. However, it is not until a second 
nick is proximal to another nick in the sister strand that degradation of the OC form is 
detected. Linear pDNA would be degraded by a mechanism similar to the degradation of 
the OC form, but also by exonucleases acting on the free ends of the plasmid. This 
additional route of degradation would contribute to L pDNA's more rapid clearance. 

Alternatively, the mechanism for the rapid clearance of SC pDNA may be due to 
physical differences between the 3 forms. Previous studies comparing SC pDNA to L 
pDNA have shown that SC pDNA has stronger acidity than L pDNA (Poly 1999). This 
difference is the result of the density and availability of the free phosphate groups. 
Acidic groups located at the external loops of SC molecules would be available and 
involved in interactions, while most of the phosphate groups localized within the SC 
molecule would not interact with components in the bloodstream. OC and L pDNA 
however likely expose a much higher number of available acidic functional groups (Poly 
1999). These anionic charges would be located all along the pDNA molecules and allow 
for multiple interactions. This decreased binding affinity of SC pDNA has been 
displayed in interactions with silica (Melzak 1996) and clay minerals (Poly 1999). The 
SC form of the plasmid has also been shown to interact more strongly with the 



100 

hydrophobic stationary phase in reversed-phase high performance hquid chromatography 
(Colote 1986). This difference in exposed electrostatic groups could potentially explain 
the rapid clearance of SC pDNA relative to OC and L pDNA. If L and OC pDNA 
interact more strongly with plasma components than SC pDNA this may decrease their 
uptake by scavenger receptors or tissues. Furthermore, this association of OC and L 
pDNA with plasma components may also offer some protection from plasma nucleases. 
Protection from nucleases has been displayed after adsorption to proteins and is the basis 
for DNase I footprinting (Lodish 1995). This would result in SC pDNA remaining free in 
the bloodstream and open to nuclease digestion. Also, the increased hydrophobicity of 
SC pDNA may also lead to a greater interaction with vascular endothelia, providing an 
additional clearance pathway, and explaining SC pDNA's larger volume of distribution. 
Thus, these physical differences may provide some insight into the observed 
pharmacokinetic differences. 

OC pDNA displayed kinetics consistent with saturable elimination. Nonlinear 
elimination of OC pDNA has previously been suggested using pharmacokinetic analysis 
of outflow patterns from rat perfused liver studies with radiolabeled OC pDNA (Yoshida 
1996). In this study Vd increased and extraction ratio decreased as perfiision dose was 
increased from 1.33 to 13.3 |iig/liver. The results presented here contribute fiirther 
evidence to support this finding. 

In conclusion, these results indicate that naked SC pDNA is cleared rapidly from 
the rat circulation after IV bolus administration at 390 (± 50) ml/min, and has a volume 
of distribution of 148 (± 26) ml. AUC analysis revealed that 60 (± 10) % of the SC 
pDNA degraded to the OC form of the plasmid. The OC form of the plasmid exhibits 



101 

nonlinear characteristics with clearance ranging from 1.3 (± 0.2) to 8.3 (± 0.8) ml/min for 
the 2500 and 250 j^g doses, respectively. Vd of the OC form was 43 (+ 15) ml. The 
conversion of the OC form of the plasmid to the L form of the plasmid appears to be 
nearly complete. The L form of the plasmid is cleared at 7.6 (± 2.4) ml/min and has a Vd 
of38(±12)ml. 



CHAPTER 5 

PHARMACOKINETIC MODELING OF PLASMID DNA AFTER IV BOLUS 

ADMINISTRATION IN THE RAT 

Introduction 

The pharmacokinetics of any drug are best studied by simuUaneous measurement 
of the parent drug and all of its pharmacologically active metabolites, especially if they 
all possess similar pharmacological properties (Garrett 1984). Plasmid DNA exists as 
three major topo forms. The native (parent) structure of non-damaged pDNA is 
supercoiled (SC). Single strand nicks in the phosphodiester backbone of the pDNA yield 
an open circular (OC) form. This metabolite of SC pDNA is still associated with 
significant activity (-90-100%) (Adami et al. 1998; Niven et al. 1998). Further single 
strand nicks to the OC pDNA yield linear (L) pDNA, associated with a significant loss of 
activity (-90%). This process is schematically illustrated in Figure 1-2. 

The mathematical description of the plasma concentration-time curve of a drug 
after administration yields only an equation describing elimination after a given dose. 
Thus, it is difficult to predict the concentration-time profile of the drug and metabolites 
after administration of varying doses. A model that quantitatively describes the 
transformation of the parent drug and its metabolites as a function of dose, may permit 
correlations with pharmacodynamic activities, and give insight into the mechanisms of 
action (Garrett 1984). 

The terms "linear" and "nonlinear" describe mathematical concepts related to a 
given plasma concentration-time dataset's dependence on administered dose (Garrett 



102 



103 

1984). Linear differential equations describe definite properties. First, transfers from 
drug to metabolite are first order. Second, multiples of dose yield the same multiple of 
drug concentration in any compartment at the same time. Finally, the plasma 
concentration-time curve can be described by a linear sum of exponentials. 

A non-linear model does not exhibit these properties (Garrett 1984). The rate of 
this type of system is not simply proportional to the plasma concentration. A well- 
characterized example of this type of system is the Michaelis-Menten model of 
elimination of a metabolized drug in a one-compartment model. This model is described 
by the following equation: 

dt K^ + C 
Where C is the parent drug's plasma concentration. Km is the concentration at which the 
system operates at half of its maximal velocity, and Vmax is the systems maximal velocity. 
In this model, drug clearance is not constant, but varies with plasma concentration. 

In order to properly dose and achieve the desired levels of protein transcript it will 
be necessary to clearly define the pharmacokinetic parameters involved with the 
clearance of pDNA from the bloodstream. Therefore it is necessary to construct a 
mathematical model to predict pDNA concentrations in vivo. We have used to the 
previous information to construct this model. Noncompartmental analysis had suggested 
that OC pDNA was subject to non-linear elimination. AUC analysis suggested that 60 
(±10) % of the SC pDNA was being converted to the OC form of the plasmid. 
Furthermore, AUC analysis also suggested complete conversion of the OC form of the 
plasmid to the L form. SC pDNA was detectable only after a 2500 j^g dose. Therefore 
the parameters for this form of the plasmid were limited to data from this dose only. 



104 

These characteristics of the noncompartmental analysis were incorporated into a model 
describing the elimination of pDNA from the bloodstream. 

The aim of this study was to accurately model the pharmacokinetics of pDNA in a 
topoform specific manner after single dose administration. This information will provide 
a basis upon which pharmacokinetic/pharmacodynamic models can be constructed. 
Furthermore, this model may provide useful in analyzing the kinetic effects of pDNA 
delivery vehicles. 

Theoretical 

SC and L pDNA concentrations were fit to pseudo first-order kinetics. OC pDNA 
was fit to Michelis-Menten kinetics. The degradation of L pDNA is considered to yield 
fragments of heterogeneous lengths, thus these products were not included in the fitted 
model. The model is presented in Figure 5-1 . Based on this model the following 
differential equafions were derived to describe the kinetics of pDNA: 

^ = -(^, + ^J.5C 

dOC . _ ^.ax(ac)OC 



= k-SC- 



dt (K„,oo + OQ 

dL^ ^.ax(OC)-QC ^^ ^ 

dt iK„,on+OQ '' 

Where SC, OC, and L are the amounts of SC, OC, and L pDNA present at time (t) 
respectively. The constant ks represents the rate constant for the degradation of SC 
pDNA to OC pDNA. The constant ku represents the removal of SC pDNA from the 
circulation or degradation to an undetectable form. The parameter Vmax(oc) is the 
apparent maximal rate of elimination of OC pDNA from the circulation. The constant 
Km(oc) represents the apparent concentration at which the kinetics operate at Yi Vmax- The 



105 



ku 



/ A 

' ■ ks Vmax(oc), Km (oc) kt 



V 



y 



Figure 5-1 . Model for pDNA clearance from the bloodstream. 



106 

constant Icl represents the first-order rate constant for L pDNA. 

The kinetics of OC pDNA after IV bolus of OC pDNA were fit to a Vmax model: 

dOC ^ VmMOQ'OC 
The kinetics of L pDNA after IV bolus of L pDNA were fit to: 

dt "^ 

Non-linear curve fitting and statistical analysis were carried out using Scientist 
(version 4.0, Micromath, Salt Lake City, UT). Goodness of fit was assesses using model 
selection criteria (MSC) (MicroMath 1995). Area under the plasma concentration time 
curve (AUC) was calculated using trapezoidal rule. Pharmacokinetic analysis was 
carried out using standard pharmacokinetic parameters as noted in the text. Statistical 
analysis was performed using SAS (Version 6.12, The SAS institute, Gary, NG). 

Results 

The concentrations of supercoiled pDNA were detectable only after the 2500 |ig 
dose. The parameters for this form of the plasmid were fixed for all other doses. 
Resulting observed and fitted data for the SC pDNA were presented in Figure 4-6. SC 
pDNA was eliminated with a ti/2 of 0.15 (± 0.01) min. 

Plasma concentrations and resulting fitted data for the OG and L pDNA are 
presented in Figure 5-2 to 5-5 for the 2500, 500, 333, and 250 ^g doses, respectively. 
Calculated pharmacokinetic parameters are presented in Table 5-1. Predicted data agreed 
well with experimental observations. Model selection criteria were 4.1, 4.3, 3.2, and 4.0 
for the 2500, 500, 333, and 250 fjg doses, respecfively. OG pDNA was eliminated with 



107 

an average Vmax of 1.7 (± 0.5) ng/^l/min and an average Km of 7.1 (±2.1) ng/)j,l over all 
doses. L pDNA was eliminated with an average ti/2 of 1 .7 (± 0.5) min. Parameters 
calculated after the data from all doses were fit simultaneously to the model are presented 
in Table 5-2. This global fitting of the data resulted in an increase in model selection 
criteria to 4.4. 

We next sought to determine if the calculated pharmacokinetic parameters would 
remain consistent after administration of each form of the plasmid independently versus 
after administration of the SC form of the plasmid. Thus OC pDNA was administered at 
2500 and 250 |u,g doses and the concentrations of OC and L pDNA in the bloodstream 
monitored. Non-linear curve fitting for the OC form of the plasmid was carried out using 
a model for saturable metabolism as described in the methods. 

Predicted concentrations of OC pDNA agreed well with experimental data. Experimental 
and observed data are presented in Figures 5-6 for the 2500 and 250 j^g doses. 
Calculated pharmacokinetic parameters are presented in Table 5-3. The elimination of 
OC pDNA was more appropriately described by Michaelis-Menten elimination with a 
statistical improvement of fit observed after utilization of this model. OC pDNA was 
eliminated with a Vmax of 1.0 (±0.3) ngVl/min and a Km of 3.9 (± 0.9) ng/^1. L pDNA 
was eliminated with an average ti/2 of 1.7 (± 0.5) min after administration of OC pDNA 
(Figure 5-7). 

The L form of the plasmid was also administered at 2500 and 250 |ig doses by IV 
bolus administration. The concentrations of L pDNA were then monitored. Non-linear 



108 







A 




















B 




























r^ 


~"~^-i-^ 
























— 101 - 

!> 






^-^ 




E 


101 - 










_^ 


" 




\ 


\ 








/T^T-. ^ 


1~"1 ^ 


1 10.- 








\l 


1 


< 

z 

Q 

Q. 


10. - 


: 








^\'' 








\ 




















; N 




1&1 






' 
















■ 







10 20 30 

time (min) 


1 

40 50 

1 


c 


10 


20 


30 

time (min) 


1 , . < 
40 


< • 1 1 > . . 

50 





Figure 5-2. Concentrations of (A) OC and (B) L pDNA in the bloodstream after 2500 j^g 
dose of SC pDNA. Data points represent the averages of n=3 ±1 standard deviation. 
Lines represent concentrations predicted by the model. 



109 



A 



B 



~ 10' 



1 10. 



10' 



/4~ 



-H 




10 15 20 

time (min) 



Iff 




10 15 20 

time (min) 



25 



Figure 5-3. Concentrations of (A) OC and (B) L pDNA in the bloodstream after 500 ^ig 
dose of SC pDNA. Data points represent the averages of n=3 ±1 standard deviation. 
Lines represent concentrations predicted by the model. 



no 



B 







, , 












101 - 










: 


■1 

c 


r**-. 


^«L 


Q 

Q. 


1CP - 
101 . 




i 


*^-^ 

( 1 - 




"^^ 



10 15 

time (min) 



20 



iw 



o 

E 

O) 

c 

< 

Z 
Q 




Figure 5-4. Concentrations of (A) OC and (B) L pDNA in the bloodstream after 333 |ig 
dose of SC pDNA. Data points represent the averages of n=3 ±1 standard deviation. 
Lines represent concentrations predicted by the model. 



Ill 





A 










B 


















(ng/mcl) 








c 


IC - 








1 10.- 


/•***~^^-^ 






< 

z 

Q 

Q. 


101 - 












^^---.^ 








yf^~^*^~* ~'^— -_ 






• 




"~~---,^ 








f 


















1 






5 


10 






5 


10 




time (min) 








time (min) 







Figure 5-5. Concentrations of (A) OC and (B) L pDNA in the bloodstream after 250 |ag 
dose of SC pDNA. Data points represent the averages of n=3 ±1 standard deviation. 
Lines represent concentrations predicted by the model. 



112 



Table 5-1. Pharmacokinetic parameters for pDNA based upon the model presented in the 
text. 



Parameter 


2500 ^g 


500 ^g 


333 ng 


250 ^ig 


ku {mm ^) 


1.9 (±0.1) 


~ 


• 


" 


^j(min'') 


2.8 (±0.2) 


~ 


~ 


- 


Vmax (ng/^l/min) 


1.3 (±0.1) 


2.0 (±0.4) 


2.1 (±0.6) 


1.3 (±0.4) 


Kn,(ng/^1) 


6.1 (±0.8) 


8.8 (±3.7) 


7.9 (±0.7) 


5.7 (±0.6) 


^/(min"') 


0.39 (±0.02) 


0.42 (±0.01) 


0.42 (±0.03) 


- 


MSC 


4.3 (±1.4) 


3.4 (±1.4) 


3.6 (±0.5) 


3.0 (±0.6) 



Parameters represent averages of n=3 ±1 standard deviation. 



113 



Table 5-2. Overall pharmacokinetic parameters for pDNA when all doses are fit 
simultaneously. 



Parameter 


Value 


Vmax (ng/|il/min) 


1.4 (±0.1) 


Kn.(ng/^1) 


7.2 (±1.9) 


^/(min"') 


0.43 (±0.16) 


MSC 


4.4 



Parameters represent simultaneously fitted values (±standard deviation of the fit) fi-om 
means of pDNA concentrations in the bloodstream after administration of SC pDNA at 
2500, 500, 333, and 250 ^g doses. 



114 

curve fitting for the L form of the plasmid was carried out using a one compartment 
model with first-order elimination as described in the methods section. 

Predicted concentrations of L pDNA agreed well with experimental data. 
Experimental and observed data for L pDNA are presented in Figures 5-8 for the 2500 
and 250 |ig doses. Calculated pharmacokinetic parameters for L pDNA are presented in 
Table 5-4. L pDNA was eliminated with a Un of 2.15 (± 0.15) min. 

Conclusions 

Noncompartmental analysis had suggested several different characteristics of 
pDNA pharmacokinetics. First it had suggested that the conversion of SC to OC pDNA 
was not a complete process. Instead 60 % of the SC pDNA appeared to be converted to 
the OC form of the plasmid. The results of the curve fitting experiments further 
supported this relationship with ks able to be fit to 60 % of the overall elimination of SC 
plasmid (i.e. the sum of ks and ku). 

A weakness of this model is that the SC form of the plasmid was detectable only 
after the 2500 |xg dose. Thus, we were forced to assume linear pharmacokinetics for this 
form of the plasmid. It is true that this elimination could occur more rapidly at lower 
doses. However, the rate at the 2500 )Lig dose is already so rapid, half-hfe = 0.15 (±0.02) 
min, that it is unlikely that fixing this parameter would significantly affect the calculated 
values for the OC and L form of the plasmid. 

Secondly, the noncompartmental analysis suggested non-linear processes were 
involved in the elimination of the OC plasmid from the bloodstream. Curve fitting to 
first order parameters also resulted in similar suggestions between doses. The value of 
the first order rate constant for OC elimination ranged from 0.035/min to 0.14/min 



115 



B 



w 



3 

■a 
< 

Z 

a 




time (min) 



10B 




10 15 

time (min) 



Figure 5-6. Concentrations of OC pDNA in the bloodstream after (A) 2500 ^ig and (B) 
250 |ig dose of OC pDNA. Data points represent the averages of n=3 ±1 standard 
deviation. Lines represent concentrations predicted by the model. 



116 



w 



IC - 



O) 



:^-*" 



< 

z 

Q 
O. 100 - 



10' 



20 



40 60 

time (min) 



80 



100 



10' 



B 



• 










y^ 


r-h-* 


*~~~*— -. 




; - 


. / 

/ 


1 




^^- 


""\ 





5 


10 


15 


20 



time (min) 



Figure 5-7. Concentrations of L pDNA in the bloodstream after (A) 2500 |a,g and (B) 250 
|a.g dose of OC pDNA. Data points represent the averages of n=3 ±1 standard deviation. 
Lines represent concentrations predicted by the model. 



117 









A 










B 
























10!. 










10!- 








3 

c 

i 

Q 

Q. 


10>. 




1 




< 

Q. 


101 - 
101 - 


•w. 












10^ 




















( 


) 2 


4 6 8 






2 4 6 8 










time (min) 








time (min) 





Figure 5-8. Concentrations of L pDNA in the bloodstream after (A) 2500 |ig and (B) 250 
(xg dose of L pDNA. Data points represent the averages of n=3 ± 1 standard deviation. 
Lines represent concentrations predicted by the model. 



118 



Table 5-3. Pharmacokinetic parameters calculated after administration of OC pDNA at 
2500 and 250 )ag doses. 



Parameter 


2500 ng Dose OC 


250 ^ig Dose OC 


Vmax (ng/|il*min) 


1.1 (±0.1) 


0.90 (± 34) 


Kn.(ng/^1) 


4.4 (± 0.4) 


3.4 (±1.1) 


ki (min"') 


0.39 (±0.12) 


0.49 (± 0.07) 


MSC 


4.4 (± 0.4) 


3.4 (±1.1) 



Parameters represent mean of n=3 ± 1 standard deviation. 



119 



Table 5-4. Pharmacokinetic parameters calculated after administration of L pDNA at 
2500 and 250 |j.g doses. 



Parameter 


2500 fig Dose L 


250 lag Dose L 


ki (min"') 


0.32 (± 0.02) 


0.33 (± 0.03) 


MSC 


2.4 (± 0.6) 


2.1 (±0.9) 



Parameters represent mean of n=3 ± 1 standard deviation. 



120 

between the 2500 and 250 \xg dose data sets, respectively. Thus, it was necessary to 
include non-linear elimination into the model to describe this change in elimination 
between doses. 

Non-linear elimination of pDNA has previously been suggested using 
pharmacokinetic analysis of outflow patterns from rat perfused liver studies with 
radiolabeled OC pDNA (Yoshida 1996). In this study volume of distribution decreased, 
as perfusion dose was increased from 1.33 to 13.3 |j,g/hver, from 0.598 (±0.09) to 0.314 
(+0.08) ml/g respectively. Extraction percentage decreased from 45.56 (±0.31) to 20.12 
(±0.75) % as dose ranged from 1.33 to 13.3 )j,g/liver respectively. Thus, the results 
presented here contribute further evidence to support nonlinear processes in the 
elimination of OC pDNA from the circulation. 

The values of Vmax and Km remained relatively constant between doses of SC 
pDNA and also after administration of OC versus after administration of SC pDNA. 
After administration of SC pDNA, the average of all doses for Vmax was 1.7 (± 0.5) 
versus 1.0 (± 0.3) ng/^l/min after administration of OC pDNA. Likewise, the value of kL 
remained relatively constant after administration of all 3 forms of the plasmid. The value 
of kL ranged from 0.47 (± 0.1 1) to 0.45 (± 0.10) to 0.32 (± 0.02) min'after administration 
of SC, OC, or L pDNA respectively. 

The presented data were all weighted with a factor of 0. This has resulted in some 
of the lower concentration data points being less well approximated by the model than 
higher concentrations. More sensitive quantitation techniques may allow the use of 
weighting factors in the fitting and provide better predictions of lower concentration data. 



121 

Thus, we conclude from the modeling experiments that SC pDNA is rapidly converted to 
the OC form of the plasmid with a half-Hfe of 0.15 (± 0.01) min. OC pDNA exhibits 
non-linear characteristics with a Vmax of 1 .5 (± 0.6) ng/nl/min and Km of 6.0 (± 2.4) 
ng/)j.l. The L form of the plasmid exhibits first-order kinetics and is eliminated with an 
overall average half-life of 1 .6 (± 0.4) min. 



CHAPTER 6 
PHARMACOKINETICS OF LIPOSOME: PLASMID DNA COMPLEXES 

Introduction 

One of the major obstacles to effective gene therapy is the generally poor 
efficiency of pDNA delivery (Thierry et al. 1997). Most gene therapy efforts involve the 
use of retroviral vectors due to their efficiency and stable integration (Thierry et al. 
1997). Clinical use of retroviral vectors, is however faced with a large number of 
obstacles. Among these are laborious preparation, difficulties in purification, concerns 
for the recombination with endogenous virus to produce a potentially infectious virion, 
and integration into the host genome to produce a tumorigenic or cytotoxic event (Liu 
1999). Because of these limitations, the use of non-viral techniques, including liposomal 
delivery, has become an intensely investigated area. An early published report displaying 
incorporation of pDNA into liposome: pDNA complexes speculated that ". . .possibly 
such liposomes could be used as vehicles for the introduction of new genes into cells" 
(Osaka et al. 1996). This report was followed by a large number of studies demonstrating 
successful in vitro liposome-mediated transgene expression in prokaryotic and eukaryotic 
cells (Osaka et al. 1996). Cationic lipids have been shown to be safe and are currently 
being utilized in gene therapy clinical trials (Valere 1999). 

The data from the previous studies presented here, display that the circulation 
time of pDNA in the circulation is remarkably short, with a rapid conversion from the SC 
form of the plasmid to the OC and L forms. This change in topoform has been shown to 
affect transcriptional activity (Murray 1991; Niven et al. 1998). For successful gene 

122 



123 

therapy, longer circulation times of the more transcriptionally active SC and OC forms of 
the plasmid would likely be beneficial. Furthermore, liposomal complexation with 
plasmid DNA offers the advantage of conjugation of a targeting ligand. Thus this may 
allow pDNA to be directed to the desired site of action. 

Although few studies are available on the pharmacokinetics of liposome:pDNA 
complexes, liposomal pharmacokinetics alone have been studied extensively with several 
reviews published (Hwang et al. 1997; Juliano 1988; Takakura et al. 1996). Liposome 
pharmacokinetics have been shown to be dependent upon size (Sato 1986), dose 
(Bosworth and Hunt 1982; Osaka et al. 1996), lipid composition (Gabizon 1988), and 
charge (Juliano 1988). In general, liposomes larger than 60 nm in diameter are unable to 
access tissues having continuous capillary endothelia, including skeletal, cardiac, and 
smooth muscle, lung, skin, subcutaneous tissue, and serous and mucous membranes, and 
are limited to uptake in tissues of the reticuloendothelial system (Hwang et al. 1997). 
Liposomes larger than 0.5 |am are confined to the vasculature in all tissues. 

After systemic administration of liposome:pDNA complexes, pDNA rapidly 
disappears from the bloodstream. Niven and coworkers (Niven et al. 1998) showed that 
2.9 % of the dose of liposome:pDNA complexes could be recovered in the bloodstream 5 
minutes after administration. Plasmid remained detectable (< 2% of the dose recovered) 
in the bloodstream through 24 h after administration. The clearance processes involve 
degradation in the blood stream (as displayed in the previous studies), interaction with 
plasma proteins, organ distribution, and uptake by the reticuloendothelial system (Juliano 
1988). The movement of liposome:pDNA complexes into tissues has been suggested to 



124 

be roughly a unidirectional system, where distribution back into the central compartment 
can be assumed to be negligible (Mahato et al. 1997). 

Previous studies had suggested that liposome: pDNA complexes offered 
protection of the pDNA from degradation by plasma nucleases (Thierry et al. 1997), but 
pDNA was removed from the circulation in a more rapid fashion than after 
adminisfration of the naked plasmid (Osaka et al. 1996; Niven et al. 1998). Niven and 
coworkers (Niven et al. 1998), using ["PJpDNA, found that 36 % of the pDNA dose 
could be recovered in the bloodstream at 5 min after administration of naked pDNA and 
only 2.9 % of the dose could be recovered at 5 min after administration of liposome: 
pDNA complexes. Osaka and coworkers (Osaka et al. 1996) also found similar results 2 
minutes after administration of liposome [^^P]pDNA complexes with 6.12 % of dose 
equivalents/g in the blood after administration of liposome:pDNA complexes versus 
15.79 % of dose equivalents/g after administration of free [^^P] pDNA. 

A large number of cationic lipids have been synthesized since the initial report in 
1987 (Liu et al. 1997). In vitro the incorporation of the neutral phospholipid 
dioleoylphosphatidylethanolamine (DOPE) into 1 ,2-dioleoyl-3-trimethylammonium 
propane (DOTAP) liposomes helps to destabilize the endocytic vacuole membrane 
allowing the release of exogenous DNA into the cytosol (Osaka et al. 1996), and 
represents a commonly used lipid mixture. While in vivo, the use of DOTAP: cholesterol 
liposomes is more common (Barron et al. 1998). Thus, these 2 lipid combinations were 
chosen to study the effects of liposome complexation on the pharmacokinetics of pDNA. 

The objective of these studies was to investigate the potential protective effects of 
liposome complexation on preservation of the SC topoform. We began by initially 



125 

studying, in vitro, the protective effects of liposome complexation in isolated plasma. 
We next analyzed the effect of liposome:pDNA ratio on this protection. Finally we 
sought to compare the effects of liposome complexation on the in vivo pharmacokinetics 
of pDNA with the pharmacokinetics of naked pDNA at equivalent dose. 

Methods 

All chemicals used were obtained similar to their description in Chapter 2 
Methods. Plasmid DNA was obtianed as described in Chapter 2 Methods. All lipids 
were purchased from Avanti Polar Lipids (Alabaster, AL). 

For in vivo experiments, liposomes were prepared by mixing l,2-dioleoyl-3- 
trimethylammonium propane (DOTAP) and cholesterol in a 1 :1 molar ratio in chloroform 
and drying the mixture under nitrogen at eCC in a Buchi Rotovapor. The dried lipid film 
was then reconstituted in sterile water containing 5% dextrose at a total lipid 
concentration of 5 mg/ml and shaken in a 60°C water bath for 15 min. Liposomes were 
then sized by sonication with an ultrasonic probe (Fisher Scientific, Springfield, NJ). 
Liposome: pDNA complexes were formed by mixing liposomes with pDNA at a 2:1 
lipid: DNA weight ratio and incubating at room temperature for 15 min. 

For in vitro experiments, lipid: pDNA complexes were formed using 1 ,2-dioleoyl- 
3-trimethylammonium propane (DOTAP) and dioleoyl phosphatidylethanolamine 
(DOPE) in a 1 :1 molar ratio. Lipid: pDNA complexes were formed by complexing 
pGE150 with DOTAP: DOPE liposomes for 15 min prior to beginning the experiment. 
Complexes were formed at 1:1, 3:1 and 6:1 lipid: DNA ratios (w/w). Complexes were 
then incubated in rat plasma and samples drawn at various time points as described 
previously. Statisfical analysis was carried out using SAS (Version 6. 12, SAS, Cary, 
NC) and a two tailed equal variance student's t-test. 



126 

For in vitro experiments, blood was isolated from male Sprague-Dawley rats 
(300-350g) by cardiac puncture, and immediately placed in heparinized test tubes 
(Vacutainer, Becton Dickinson, Franklin Lakes, NJ) on ice. Blood samples were 
centrifuged at 6,000 g for 5 min. 600 |al of plasma was removed and placed on ice until 
assay. Plasma samples were warmed to 37° C in a water bath, and maintained at 37° C 
for the duration of the experiment. 12 i^g of pDNA, in phosphate buffered saline, was 
incubated in the 37° C plasma and 50 |j,l samples were taken at the times indicated. 80 |j.l 
of phenol: chloroform: isoamyl alcohol (25: 24: 1, v/v/v) was immediately added to each 
sample, vortexed for 5 s at low speed, and placed on ice. Samples were centrifuged at 
20,800 g for 10 min at room temperature. From the supernatant, an aliquot of 15 |al was 
removed, 5 lal of 1 x loading dye (Promega, Madison, WI) added, and placed on ice until 
loaded on an agarose gel. A final volume of 10 ^1 was loaded on agarose gels. Analysis 
and quantitation was performed as described in Chapter 2 Methods. 

IV administration and analysis of liposome: pDNA complexes were performed as 
described earlier in Chapter 3 methods. Briefly, SC liposome: SC pDNA complexes 
were administered to male Sprague-Dawley rats (300-350 g) by IV bolus in the femoral 
vein at a dose equivalent to 500 |ag of pDNA. Blood samples were drawn through a 
jugular vein cannula. Isolated blood samples (approximately 300 i^l) were immediately 
placed in tubes containing 57 |il of 0.34mM EDTA. Samples were then liquid: liquid 
extracted with phenol: chloroform: isoamyl alcohol (25: 24:1 v/v/v). Quantitative 
analysis was performed as described earlier in Chapter 3 methods. 



127 

Results 
Pharmacokinetics of Liposome:pDNA Complex Degradation in Rat Plasma 

To investigate the potential protective effects of cationic liposomes from plasma 
nucleases, we complexed a common liposomal delivery vehicle, DOTAP: DOPE (1:1 
w/w), to pDNA (3:1 lipid: DNA, w/w) and incubated the complexes in freshly isolated rat 
plasma at 37°C. Figure 6-1 displays that a portion of the SC pDNA remained detectable 
through 5.5 h (8.2% of the time=0 amount). 

We next sought to determine the effect of lipid: pDNA ratio on the degradation 
observed in the previous experiment. We hypothesized that increasing the lipid:pDNA 
ratio should offer increased protection from nucleases. We chose to investigate this by 
complexing lipid: pDNA at 1 : 1 . 3 : 1 , and 6: 1 ratios. Agarose gel analysis revealed that 
the SC pDNA was detectable through at least 5 hours in all three ratios (Figure 6-2). 
Quantification of the percent SC remaining revealed that 28.9%, 37.8%, and 17.7% (for 
6:1, 3:1, and 1:1 respectively) of the 1 minute amount remained at 3 hours. A statistical 
analysis of the percent SC plasmid remaining at revealed that there was no statistically 
significant difference between the percent remaining in the 6: 1 versus 1:1,3:1 versus 1 : 1 
and the 6:1 versus 3:1 ratios. These results suggest increasing the lipid: DNA ratio from 
1:1 to 6:1 offers no significant increase in protection from plasma nucleases through 3 
hours. 

The OC and L forms of the plasmid were hypothesized to appear for three 
possible reasons. First, it is possible that some portion of the pDNA remains non- 
complexed and free in solution. When this mixture of complexed pDNA and free pDNA 
is incubated in the plasma, the free pDNA degrades as previously described, and thus the 
OC and L pDNA appear and degrade. A second possible explanation is that the pDNA 



128 

on the outer surface of the Uposome: pDNA complex aggregates is offered Uttle 
protection from plasma nucleases and is degraded. The pDNA towards the core of the 
complex is then spared from degradation. Finally, it is possible that the complexes 
degrade in the plasma. 

To confirm that all pDNA was complexed, we complexed lipid: pDNA at 1:1, 3:1, 
and 6:1 ratios (w/w) and analyzed their migration through an agarose gel. If any pDNA 
remained non-complexed, it should migrate through the gel and separate from the 
complexed pDNA remaining in the well. No migration from the wells was observed at 
any lipid: pDNA ratio as seen in Figure 6-3. These results suggest that all pDNA is 
complexed at all three ratios, pDNA must therefore be degrading from the surface of the 
liposomes or the complexes must dissociate in the plasma. 
Pharmacokinetics of Liposome:pDNA Complex Degradation in Rat Whole Blood 

We next sought to determine if the effects we observed using liposomes in the 
plasma were consistent with the effects we would observe in whole blood. Unlike naked 
pDNA, liposome/pDNA complexes have been shown to be recovered in the red blood 
cells (37 to 84 %) (Osaka et al. 1996). We hypothesized that this uptake may affect the 
kinetics observed. Liposome/pDNA complexes taken up by the red blood cells may be 
may be disassembled in the RBC and pDNA degraded when freed from the protective 
liposome complex. Thus our hypothesis was that pDNA would break down faster in 
whole blood. We also sought to determine the effects of increasing the lipid:pDNA ratio 
on this process, and secondarily hypothesized that increasing the lipid:pDNA ratio should 
lead to more rapid uptake by RBC, resulting in more rapid degradation of the pDNA. 

The results of the whole blood experiments revealed that the protocol used in 
isolation of pDNA from the plasma also were sufficient to isolate pDNA from whole 



129 

blood. Plasmid DNA incubated in whole blood did not degrade faster than pDNA in the 

plasma (Figure 6-4 A). Furthermore, increasing the lipid:pDNA ratio from 3:1 to 6:1 did 

not serve to increase the degradation observed (Figure 6-4 B). 

These results suggest that the presence of RBC does not significantly affect the 

degradation rate observed. An increase in lipid:pDNA ratio also does not affect the 

degradation observed in the plasma. Liposome complexation results in a relative 

maintenance of the SC topoform with 28.9%, 37.8%, and 17.7% (for 6:1, 3:1, and 1:1 

respectively) of the 1 minute amount remaining after 3 hours. 

Pharmacokinetics of Liposome:pDNA Complexes after IV Bolus Administration in the 
Rat 

The in vitro results displayed that liposome complexation can protect a portion of 
the SC topoform from degradation through 5.5 hours. We next sought to determine if 
liposome complexation would provide similar protection in vivo. Thus, we injected 
liposome: SC pDNA complexes at a 500 ^ig dose of pDNA as described in the methods 
section. 

Agarose gel analysis obtained after administration of liposome: pDNA complexes 
is presented in Figure 6-5. Plasma concentrations of SC, OC, and L pDNA are presented 
in Figure 6-6. Unlike after administration of naked pDNA, SC pDNA was readily 
detectable after administration, and remained detectable through 5 minutes after 
administration of the 500 i^g dose. However, the OC and L form of the plasmid also only 
remained detectable through 5 min after administration, versus through 20 minutes after 
administration of naked pDNA. 

Noncompartmental analysis of all three forms of the plasmid is presented in Table 
6-1, and a relative comparison of the naked and liposome complex parameters is 



130 

presented in Tables 6-2, 6-3, and 6-4 for SC, OC, and L pDNA, respectively. If we again 
make the assumption that SC pDNA follows linear pharmacokinetics, mean residence 
time for the SC form of the plasmid increased from 0.21 (± 0.02) after administration of 
naked pDNA to 0.99 (± 0.42) min after administration of liposome:SC pDNA complexes. 
Clearance decreased from 390 (± 10) to 87 (± 30) ml/min. Obviously this disparity 
would only become more severe if the true pharmacokinetics of naked SC pDNA are 
more rapid than this at the 500 )ag dose. Area under the concentration time curve of the 
OC form of the plasmid decreased after liposome complexation from 120 (±50) to 14 
(±4) ng/|al*min. Clearance/f of the OC form of the plasmid increased from 6.9 (±2.8) to 
95 (±37) ml/min. Area under the concentration time curve (AUC) of L pDNA also 
decreased from 52 (±25) to 5.7 (±1.9) ng/(il*min. 

Conclusions 

These results indicate that liposome complexation can indeed offer protection of 
the SC form of the plasmid. This increase in circulation half-life provides fiirther 
evidence to support the use of liposomal delivery vehicles in gene therapy. This analysis 
also provides evidence that, although the liposome complexation does protect the SC 
form of the plasmid, the individual particles have an overall more rapid clearance from 
the circulation. The OC and L form of the plasmid also only remained detectable through 
5 min after administration, versus through 20 minutes after adminisfration of naked 
pDNA. 

This finding was mirrored by the ["P]DNA results of Niven and coworkers 
(Niven et al. 1998) who found that 36 % of the pDNA dose could be recovered in the 
bloodstream at 5 min after administration of naked pDNA and only 2.9 % of the dose 



131 



^^"^ — !^ -^ — ■"" 



'^'&^m$4i^^ 




Figure 6-1 . Liposome-pDNA complexes were incubated in rat plasma for various time 
points. 10 |il of sample was loaded in each lane as described in the methods section. 
Lane 1; size standard, lane 2; 1 min, lane 3; 2 min, lane 4; 5 min, lane 5; 10 min, lane 6; 
20 min, laneV; 30 min, lane 8; 60 min, lane9; 2 h, lane 10; 3 h, lane 1 1; 5.5 h. 



132 



B 




^SC 



<-sc 



<-sc 



Figure 6-2. Agarose gel analysis of liposome/pDNA complexes. (A) 1:1 lipid:pDNA 
ratio, through 4 hours. (B) 3:1 lipid:pDNA ratio, through 6 hours. (C) 6:1 lipid:pDNA 
ratio, through 6 hours. *Indicates the 3 hour time point. 



133 




3:1 6:1 



Figure 6-3. Lane 1: high molecular weight size standard, lane 2: 1:1 lipid:pDNA 
complexes, lane 3:3:1 Iipid:pDNA ratio (w/w), lane 4: 6:1 lipid:pDNA ratio 



134 



B 





50 



) 10 


15 



time 





(min) 






.3to1 
.6to1 



50 



time 
(min) 



10 




15 




Figure 6-4. (A)Degradation of SC pDNA in rat blood versus plasma. (B)Degradation of 
supercoiled pDNA in 3: 1 and 6: 1 (w/w) liposome/pDNA complexes incubated in 
heparinized rat whole blood. Error bars indicate standard deviation of n=3 rats. 



135 



10 




Figure 6-5. Agarose gel analysis of pDNA after administration of liposome: pDNA 
complexes. Lane 1: 15 sec, lane 2: 30 sec, lane 3: 45 sec, lane 4: 1 min, lane 5: 1.5 min, 
lane 6: 2 min, lane 7: 2.5 min, lane 8: 3 min, lane 9: 4 min, lane 10: 5 min. 



136 



10 



S 1 



Q 

a 



0.1 








2 3 

time (min) 



Figure 6-6. Plasma concentrations of SC, OC, and L pDNA after 500 |ig IV bolus 
administration of SC pDNA: liposome complexes. Key: ♦: SC, ■: OC, A: L. 



137 



Table 6-1. Noncompartmental analysis of pDNA after administration of liposome: SC 
pDNA complexes. 



Parameter 


Supercoiled 


Open Circular 


Linear 


AUC (ng/nl*min) 


0.0063 (± 0.0023) 


0.014 (±0.004) 


0.0057 (±0.0019) 


AUC % 
extrapolated 


12 (±3) 


25 (±18) 


68 (± 26) 


AUMC 

(ng/^l*min^) 


0.0068 (± 0.0053) 


0.04 (± 0.004) 


0.017 (±0.003) 


MRT (min) 


0.99 (± 0.42) 


3.0 (± 0.9) 


3.3 (±1.0) 


Cl/f(mymin) 


87 (± 30) 


37 (±9) 


95 (± 37) 



Parameters represent averages of n=3 ± 1 standard deviation. 



138 



Table 6-2. Comparison of SC pDNA pharmacokinetic parameters after administration of 
SC pDNA either in free form (naked) at 2500 |a,g dose or after administration as 
liposome: pDNA complexes at 500 \ig dose. 



Parameter 


Naked SC pDNA 


Liposome: SC pDNA 
complexes 


MRT (min) 


0.21 (+0.02) 


0.99 (± 0.42) 


CI (ml/min) 


390 (±10) 


87 (± 30) 


Vdss (ml) 


148 (±26) 


79 (± 16) 



Parameters represent averages of n=3 ± 1 standard deviation. 



139 



Table 6-3. Comparison of OC pDNA pharmacokinetic parameters after administration of 
SC pDNA either in free form (naked) or after administration as liposome: pDNA 
complexes at 500 |j,g pDNA dose. 



Parameter 


OC pDNA after Naked SC 
pDNA 


OC pDNA after 

LiposomerSC pDNA 

Complexes 


AUC (ng/nl*min) 


120 (±50) 


14 (±4) 


AUC % extrapolated 


9 (±4) 


25 (±18) 


AUMC (ng/nl*min^) 


1900 (±1200) 


40 (± 4) 


MRT (min) 


15 (±4) 


3.0 (± 0.9) 


Cl/f(mymin) 


3.0 (±1.2) 


37 (± 9) 


Cmax (ng/)4,l) 


13 (±4) 


5.8 (± 0.7) 


Tmax (min) 


1(±0) 


0.25 (±0) 



Parameters represent averages of n=3 ± 1 standard deviation. 



140 



Table 6-4. Comparison of L pDNA pharmacokinetic parameters after administration of 
SC pDNA either in free form (naked) or after administration as liposome: pDNA 
complexes at 500 |j.g pDNA dose. 



Parameter 


L pDNA after naked SC 
pDNA 


L pDNA after Liposome: 
SC pDNA complexes 


AUC (ng/|al*min) 


52 (±25) 


5.7 (±1.9) 


AUC % extrapolated 


15 (±5) 


68 (± 26) 


AUMC (ng/^l*min') 


570 (±370) 


17 (±3) 


MRT (min) 


10 (±2) 


3.3 (±1.0) 


Cl/f(ml/min) 


6.9 (±2.8) 


95 (± 37) 


Cmax (ng/|al) 


3.2 (±1.0) 


1.8 (±0.8) 


Tmax (min) 


5.3 (±4.0) 


0.25 (±0) 



Parameters represent averages of n=3 (± 1 standard deviation). 



141 



could be recovered at 5 min after administration of liposome: DNA complexes. Osaka 
and coworkers (Osaka et al. 1996) also found similar results 2 minutes after 
administration of liposome ["P]DNA complexes with 6.12 % of dose equivalents/g in the 
blood versus 15.79 % of dose equivalents/g after administration of free [^^P] DNA. 

Thierry and coworkers (Thierry et al. 1997) utilized lipospermine (DOGS) and 
DOPE liposomes and also were able to display protection of pDNA from degradation. In 
this study SC pDNA was detected for up to 60 min after incubation in isolated plasma, 
but was rapidly eliminated when IV delivered. Thierry estimated the plasma half-life of 
the OC pDNA to be much longer than the results presented here at 10 to 20 minutes. 
Their work did however, show for the first time, the presence of intact pDNA in plasma 
and blood cells following systemic administration. Lew and coworkers (Lew et al. 1995) 
used l,2-dimyristoyl-oxypropyl-3 -dimethyl ammonium bromide: DOPE liposomes and 
determined a half-life of a few minutes for the OC form of the pDNA, similar to the 
results presented here. However, unlike the results presented here, they reported no 
detectable SC pDNA. It may be that cationic liposome: pDNA complexes may bind to 
serum components, and these interactions may differ with different lipids. Indeed, 
altering lipids and lipid to lipid weight ratios does affect fransfection levels of various 
organs in vivo (Liu 1997; Song et al. 1997). 

Liu and coworkers (Liu et al. 1997) showed that higher cationic lipid to pDNA 
ratio was essential to achieve better gene delivery efficiency in vivo. The results 
presented in our studies, however, displayed that increasing the lipid to pDNA ratio from 
1:1 to 6:1 had no effect on the observed protection from plasma nucleases in isolated 



142 

plasma or whole blood. However, analysis of Liu and coworkers results display that 
transfection activity remained relatively stable between the 1:1 and 6:1 ratios. Ratios 
higher than 24:1 were essential to achieve significant increases in transfection activity. 
Thus, even higher ratios of lipid: pDNA may provide protection from plasma nucleases. 

Alternatively, one must also consider the dilution effect in the bloodstream after 
administration of such a necessarily large volume at these higher lipid: pDNA ratios. It is 
likely that mouse vasculature is occluded with these high volumes. Thus, complexes 
travel in the vasculature with little mixing with blood. This would obviously provide 
protection of the pDNA. However, given that these volumes on a weight basis would not 
be used in clinical trails in humans (Valere 1999), it is unlikely that this vascular 
occlusion effect would be so pronounced. 

Furthermore, there is likely a maximal amount of complex that can be taken up by 
tissues. If there is a maximal amount that can be taken up, then there would be a 
saturation lipid dose that, above which, no further increases in transfection would be 
observed. This, in fact, was observed in Song and coworkers (Song et al. 1997) and Liu 
and coworkers (Liu et al. 1 997) experiments. Song and coworkers found that ratios from 
36:1 to 48:1 offered no increase in transfection, while increases from 2:1 to 36:1 offered 
significant increases. Liu and coworkers found that ratios between 24:1 and 48:1 also 
offered no increase in transfection, while increasing the ratio from 6:1 to 24:1 offered 
significant increases. Thus, these extremely high ratios of lipid: pDNA may be 
unnecessary to provide adequate protection of the pDNA. 

The OC and L pDNA were cleared from the circulation more rapidly than after 
administration as naked pDNA. This may be due to the fact that liposome pDNA 



143 

complexes are more rapidly cleared from the circulation than naked pDNA, regardless of 
the pDNA form. If this were true, one would predict that administration of radiolabeled 
pDNA as liposome: pDNA complexes should show higher and more rapid increases of 
tissue radioactivity than after administration as naked pDNA. This results has been 
observed after administration of ["P] pDNA (Osaka et al. 1996; Liu 1997; Song et al. 
1997; Niven et al. 1998). This was evident as increases in AUC maximal radioactivity 
recovery/ tissue weight. Blood radioactivity exhibited larger AUCs' and CmaxS' after 
administration of naked pDNA. This provides further evidence to support this 
hypothesis. 

In conclusion, liposome pDNA complexes are eliminated from the circulation 
more rapidly than naked pDNA, while providing protection of the SC topoform from 
degradation in the bloodstream. This level of protection is independent of lipid: pDNA 
ratio from 1:1 through 6:1. SC pDNA is detectable through 5 min after administration as 
liposome: pDNA complexes at a 500 |ag dose, whereas it is undetectable after 
administration as naked pDNA at this dose. Clearance of the SC pDNA decreased from 
390 (± 10) to 87 (± 30) after administration as naked or liposome complexes, 
respectively. Volume of distribution decreased from 148 (± 26) to 79 (± 16) ml after 
adminisfration as naked or liposome complexes, respectively. OC and L pDNA exhibited 
decreases in Cmax after administration as liposome: SC pDNA complexes from 13 (± 4) to 
5.8 (+ 0.7) ng/|^l and 3.2 (± 1 .0) to 1 .8 (± 0.8) ng/^1, respectively. Decreases in tmax after 
administration as liposome: SC pDNA complexes were also displayed for both OC and L 
pDNA from 1 (±0) to 0.25(±0) and 5.3 (±4.0) to 0.25 (± 0) min, respectively. Clearance/f 
increased after administration as liposome: SC pDNA complexes for both the OC and L 



144 

forms of the plasmid from 3.0 (± 1 .2) to 37 (± 9) and 6.9 (± 2.8) to 95 (± 37) ml/min, 
respectively. Thus, liposome pDNA complexes are eliminated from the circulation more 
rapidly than the naked pDNA, while providing protection of the SC topoform from 
degradation in the bloodstream. 



CHAPTER 7 
CONCLUSIONS AND IMPLICATIONS 

Summary of Results 

Implications of Plasmid DNA Degradation in Isolated Plasma 

DNase I is a well-characterized enzyme in human plasma present at 

concentrations averaging 26.1 (±9.2) ng/ml in the sera of healthy humans (Chitrabamrung 
1981). Traditionally the presence of this enzyme has led to the conclusion that pDNA 
administered IV is degraded in a rapid fashion (Gosse et al. 1965; Chused 1972). This 
has led to the current view of gene delivery, in which protection from plasma nucleases is 
a major goal of delivery vehicles. The results of this study reveal that although the half- 
life of SC and OC pDNA is remarkably short, degradation alone was not enough to 
explain the rapid disappearance of pDNA from the circulation observed in vivo. After IV 
bolus the rate of degradation of SC pDNA was greater than 7 times faster than in isolated 
plasma. 

Previous reports on the pharmacokinetics of pDNA have only been qualitative, or 
involved radiolabeling. These studies indicated that pDNA degrades within 5 minutes 
after incubation in whole blood in vitro or after IV injection in mice (Kawabata et al. 
1995; Thierry et al. 1997). We sought to quantitatively model the pharmacokinetics 
imderlying the stability of pDNA in the plasma using isolated rat plasma as a model 
system. 

The results presented in Chapter 2 revealed that SC pDNA degrades in isolated 
plasma with a half-life of 1 .2 min. Open circular pDNA is more stable than the 

145 



146 

supercoiled topoform degrading with a half-life of 21 min. Linear pDNA is degraded 
more rapidly than the OC topoform with a half-life of 1 1 min. A schematic 
representation of the kinetics of pDNA degradation in isolated rat plasma is presented in 
Figure 7-1. 

Liposome complexation revealed a relative maintenance of the SC topoform 
through 5.5 h. This provides further evidence to suggest that liposome complexation may 
not only be a means by which to deliver pDNA to target sights, but also to specifically 
protect SC pDNA from degradation. 

In summary the in vitro work presents a pharmacokinetic model describing the 
degradation of pDNA in rat plasma. Using the model derived, we are able to conclude 
that naked supercoiled pDNA degrades in rat plasma with a half-life of 1.2 (± 0.1) min, 
open circular with a half-life of 21 (± 1) min, and linear pDNA with a half-life of 1 1 (± 2) 
min. Furthermore, these studies provide evidence that supercoiled pDNA can remain 
stable in the plasma through 5.5 hours when complexed to cationic liposomes. This 
degradation was independent of sequence between the pGL3, pGE150 and pGeneMax- 
Luciferase plasmids. 
Comparison of In Vitro and In Vivo Pharmacokinetics 

The results presented in Chapter 3 indicate that SC pDNA was undetectable after 
IV bolus administration of a 500 |ag dose, whereas SC pDNA was readily detectable in 
isolated plasma, and remained detectable through 3 min of incubation. Similar results 
were seen for the OC and L forms of the plasmid. The terminal half-lives of OC and L 
pDNA decreased from 21 (±1) to 5.3 (±1.4) and 1 1 (±2) to 1.9 (±0.8) min, respectively. 
This indicates that nuclease activity alone is not sufficient to describe the rapid clearance. 



147 




0000 



ks: 0.59 
(±0.03) min' 



sc 



i 

O 
i 



oc 



ko: 0.033 
(±0.002) min 




Figure 7-1 . Schematic representation oFpDNA degradation in isolated plasma 



148 

of pDNA from the bloodstream in rats 

Chused and coworkers (Chused 1972) also suggested that nuclease activity was 
not enough to explain the rapid clearance of KB cell DNA from the circulation in mice. 
In this study, only 2 to 3 % of the radioactivity was hydrolyzed to trichloroacetic acid 
(TCA) soluble fragments in 30 min, which was several half-Hves of the DNA in the 
circulation. Tsumita and Iwanga (Tsumita and Iwanga 1963) also found that less than 5 
% of the total radioactivity was found in the TCA soluble fraction after 4.5 hours in 
mouse serum. 

Alternatively, Gosse and coworkers (Gosse et al. 1965) suggested a major role for 
nucleases in the initial degradation of DNA after IV administration in rabbits and mice. 
This finding was based upon the proportionality between the initial rate of 
depolymerization and the plasma DNase activity level. Also, a rapid decrease in 
viscosity of isolated blood was discovered indicating a depolymerization of DNA. 
Finally a markedly slower disappearance of DNA-methyl green complex (a non-specific 
DNase inhibitor) than after native DNA. 

The reason for this disparity in results deserved fiirther investigation. Gosse 
utilized much higher doses of pDNA in their invesfigafions, 200 i^g versus 5 )a.g pDNA in 
Chused and coworkers 's investigations. This disparity may be due to saturation of a 
scavenger receptor, allowing nuclease activity to become increasingly important. The 
effect of increasing dose on the clearance of DNA deserved fiarther investigation. 
Effects of Increasing Dose of Plasmid DNA 

The results presented in Chapter 4 reveal that all forms of pDNA (SC, OC, and L) 
are rapidly cleared from the circulation. Other investigators have qualitatively 



149 

commented on the rapid clearance observed after IV bolus administration of pDNA (Lew 
et al. 1995; Mahato et al. 1997; Thierry et al. 1997). However, these studies have been 
Hmited to the OC and L forms of the plasmid. The half-life of the SC topoform has been 
unable to be estimated due to lack of detection (Lew et al. 1995; Thierry et al. 1997). 

The SC form of the plasmid was detectable only after the 2500 |ag dose, and 
disappeared from the circulation after 1 min. One possible explanation for the rapid 
disappearance of SC pDNA is that it is rapidly being converted to the OC form in vivo by 
endogenous nucleases present in the plasma. If we compare the reported concentrations 
of DNase I in human plasma along with the reported SC pDNA nicking activity of DNase 
I under optimal conditions (Dwyer 1999) and make the assumption that the activity of rat 
DNase 1 is approximately the same as the human isoform (Takeshita et al. 1996), we can 
arrive at an approximate activity of 0.1 ng pDNA/|u,l/min. This is far less than the in vivo 
SC pDNA nicking rate of 9.2 ng pDNA/^l/min observed 30 sec after administration of 
the 2500 |a,g dose. One minute after administration the in vivo rate was 2.7 ng 
pDNA/|^l/min. Thus, the activity of DNase I would not seem to be enough to describe 
the rapid conversion of SC pDNA to the OC form. The combined effect of enzymes in 
addition to DNase I is also insufficient to describe the kinetics in total. After IV bolus 
administration, the rate of degradation of SC pDNA was greater than 7 times faster than 
in isolated rat plasma. 

The mechanism for the rapid clearance of SC pDNA may also be due to physical 
differences between the 3 forms. Previous studies comparing SC pDNA to L pDNA have 
shown that SC pDNA has stronger acidity in solution than L pDNA (Poly 1999). This 
difference is the result of the density and availability of the free phosphate groups. 



150 

Acidic phosphate groups located at the external loops of SC molecules would be 
available and involved in interactions, while most of the phosphate groups localized 
within the SC molecule would not interact with components in the bloodstream. OC and 
L pDNA however likely expose a much higher number of available acidic functional 
groups (Poly 1999). These anionic charges would be located all along the pDNA 
molecules and allow for multiple interactions. This decreased binding affinity of SC 
pDNA has been displayed in interactions with silica (Melzak 1996) and clay minerals 
(Poly 1999). The SC form of the plasmid has also been shown to interact more strongly 
with the hydrophobic stationary phase in reversed-phase high performance liquid 
chromatography (Colote 1986). This difference in exposed electrostatic groups could 
potentially explain the rapid clearance of SC pDNA relative to OC and L pDNA. If L 
and OC pDNA interact more strongly with plasma components than SC pDNA this may 
decrease their uptake by scavenger receptors or tissues. Furthermore, this association of 
OC and L pDNA with plasma components may also offer some protection from plasma 
nucleases. Protection from nucleases has been displayed after adsorption to proteins and 
is the basis for DNase I footprinting (Lodish 1995). This would result in SC pDNA 
remaining free in the bloodstream and open to nuclease digestion. Also, the increased 
hydrophobicity of SC pDNA may also lead to a greater interaction with vascular 
endothelia, providing an additional clearance pathway, and explaining SC pDNA's larger 
volume of distribution. Thus, these physical differences may provide some insight into 
the observed pharmacokinetic differences. 

The OC pDNA displayed kinetics consistent with saturable elimination. Curve 
fitting to first order parameters also resulted in similar suggestions between doses. The 



151 

value of the first order rate constant for OC elimination ranged from 0.035/min to 
0.14/min between the 2500 and 250 )ig dose data sets, respectively. Thus it was 
necessary to include non-linear elimination into the model to describe this change in 
elimination between doses. 

In conclusion, these results indicate that naked SC pDNA is cleared rapidly from 
the rat circulation after IV bolus administration at 390 (± 50) ml/min, and has a volume 
of distribution of 148 (± 26) ml. AUC analysis revealed that 60 (± 10) % of the SC 
pDNA appeared as the OC form of the plasmid. The OC form of the plasmid exhibits 
nonlinear characteristics with clearance ranging fi-om 1.3 (± 0.2) to 8.3 (± 0.8) ml/min for 
the 2500 and 250 ^g doses, respectively. Volume of distribution of the OC form was 43 
(± 15) ml. The conversion of the OC form of the plasmid to the L form of the plasmid 
appears to be nearly complete. The L form of the plasmid is cleared at 7.6 (± 2.4) ml/min 
and has a volume of distribution of 38 (± 12) ml. 
Results of the Curve Fitting Experiments 

The model presented in Figure 5-1 successftilly described the data. The values of 
Vmax and Km for OC pDNA remained relatively constant between doses of SC pDNA and 
also after administration of OC versus after administration of SC pDNA. After 
administration of SC pDNA, the average of all doses for V^ax was 1.7 (± 0.5) versus 1.0 
(± 0.3) ng/|^l/min after administration of OC pDNA. Likewise, the value of kt remained 
relatively constant after administration of all 3 forms of the plasmid. The value of kt 
ranged from 0.47 (± 0.1 1) to 0.45 (± 0.10) to 0.32 (± 0.02) min"' after administration of 
SC, OC, or L pDNA respectively. 



152 

Thus, we conclude from the modeling experiments that SC pDNA is rapidly 
converted to the OC form of the plasmid with a half-life of 0.15 (± 0.01) min. OC pDNA 
exhibits non-linear characteristics with a Vmax of 1 .5 (± 0.6) ng/|al/min and K^ of 6.0 (± 
2.4) ng/)j,l. The L form of the plasmid exhibits first-order kinetics and is eliminated with 
an overall average half-life of 1.6 (+ 0.4) min. Schematic representationof pDNA 
pharmacokinetic parameters after IV bolus administration of SC pDNA is presented in 
Figure 7-2. 

Liposome: pDNA Complex Conclusions 

These results indicate liposome complexation can indeed offer protection of the 
SC form of the plasmid. This increase in circulation half-life provides further evidence to 
support the use of liposomal delivery vehicles in gene therapy. This analysis also 
provides evidence that, although the liposome complexation does protect the SC form of 
the plasmid, the individual particles have an overall more rapid clearance from the 
circulation. The OC and L form of the plasmid also only remained detectable through 5 
min after administration, versus through 20 minutes after administration of naked pDNA. 

Liposome pDNA complexes were eliminated from the circulation more rapidly 
than the naked pDNA, while providing protection of the SC topoform from degradation 
in the bloodstream. This level of protection is independent of lipid: pDNA ratio from 1:1 
through 6:1. SC pDNA is detectable through 5 min after administration as liposome: 
pDNA complexes at a 500 |.ig dose, whereas it is undetectable after administration as 
naked pDNA at this dose. Clearance of the SC pDNA decreased from 390 (± 10) to 87 



153 

(± 30) after administration as naked or liposome complexes, respectively. Volimie of 
distribution decreased from 148 (± 26) to 79 (± 16) ml after administration as naked or 



154 




1.9 (± 0.1)min ' 

Vmax 1.5 (±0.6) ng/|il/min 



Figure 7-2. Schematic representation of pDNA pharmacokinetic parameters after IV 
bolus administration of SC pDNA in the rat. 



155 

liposome complexes, respectively. OC and L pDNA exhibited decreases in Cmax after 
administration as liposome: SC pDNA complexes from 13 (± 4) to 5.8 (± 0.7) ng/fj,l and 
3.2 (± 1 .0) to 1 .8 (± 0.8) ng/|_il, respectively. Decreases in tmax after administration as 
liposome: SC pDNA complexes were also displayed for both OC and L pDNA from 1 
(±0) to 0.25(±0) and 5.3 (±4.0) to 0.25 (± 0) min, respectively. Clearance/f increased 
after administration as liposome: SC pDNA complexes for both the OC and L forms of 
the plasmid from 3.0 (± 1.2) to 37 (± 9) and 6.9 (± 2.8) to 95 (± 37) ml/min, respectively. 

These results indicate that liposome pDNA complexes are eliminated from the 
circulation more rapidly than the naked pDNA, while providing protection of the SC 
topoform from degradation in the bloodstream. This theory is schematically represented 
in Figure 7-3. This results presented here are consistent with previously published results 
(Osaka et al. 1996; Niven et al. 1998). Thus, liposome complexation may be an atfractive 
means by which to protect the SC topoform, but the complex ability to target specific 
organs may be limited by their rapid clearance from the circulation. 

Future Directions 

The results and models presented here successfiiUy described the 
pharamcokinetics of pDNA after IV administration in the rat. However, there are still a 
large number of studies that need to be performed. Among these directions, an important 
area is to link the PK model presented here to pharmacodynamic (PD) effect. Selection 
of an appropriate PD effect is not as direct of a relationship as for traditional drugs 
exhibiting receptor mediated effects. 



156 




Figure 7-3. Schematic representation of liposome pDNA clearance from the 
bloodstream. In this model, removal from the bloodstream of the lipid: pDNA complexes 
is assumed to be larger than the degradation of the complex. 



157 



The most direct mesure of the dose-response-over time relationship between 
pDNA and PD may be to measue the levels of mRNA transcript. This analysis could be 
performed by quantitative polymerase chain reaction. Using this method, levels of 
transcribed mRNA in various tissues could be calcualted over time and correlated to the 
PK model presented here. 

Alternatively, one could model the levels of transcribed protein in tissues over 
time as a PD parameter. This could be done by a number of methods including enzyme 
linked immimosorbent assay, western blotting, radio immuno assay, or 
immunoprecipitation. This measurement is a more direct relationship to clinical 
response. However, this relationship may be more difficult to obtain given that the 
translation into a final protein is a multistep process. It is likely that pDNA, mRNA, and 
protein stability in the cytosol differs between cells of various tissues and between cell 
types of a given tissue. Thus, correlation of protein levels may be more difficult to 
achieve. The kinetic processes within a cell will be an important area to be understood, 
in addition to kinetic processes in the bloodstream. 

Furthermore, the application of this model may also need to be adjusted for 
different species. The clearance of DNA has been shown to be similar between strains of 
New Zealand, DBA/2, and BALB/c mice (Chused 1972), but differed between mice and 
rabbits (Gosse et al. 1965). Thus, this model may need to be adjusted not only for 
parameter values, but also in necessary parameters to describe the data. 

Concluding Remarks 

In conclusion, this work presents a pharmacokinetic analysis of pDNA in vitro, in 
isolated rat plasma, and in vivo, after IV bolus administration in the rat. Comparison of 



158 

the in vitro and and in vivo results displayed that degradation by plasma nucleases was 
not sufficient to describe the pharmacokinetics of pDNA. In addition, the 
pharmacokinetic effects of liposome complexation were investigated in vitro and in vivo. 
These results displayed that liposome complexation was a means by which the SC 
topoform could be protected, but the complexes had an overall more rapid clearance from 
the bloodstream in vivo. 



LIST OF REFERENCES 

Adami, R. C, Collard, W.T., Gupta, S.A., Kwok, K.Y., Bonadio, J., and Rice, K.G. 
(1998). "Stability Of Peptide-Condensed Plasmid DNA Formulations." J Pharm Sci 
87(6): 678-683. 

Barron, L. G., Meyer, K.B.. and Szoka, F.C. (1998). "Effects Of Complement Depletion 
On The Pharmacokinetics and Gene Delivery Mediated By Cationic Lipid-DNA 
Complexes." Hum Gene Ther 9: 315-323. 

Bosworth, M. E., And Hunt, C.A. (1982). "Liposome Disposition In Vivo II: Dose 
Dependency." PharmRes 71(1): 100-104. 

Bright, R. K., B. Beames, M. H. Shearer and R. C. Kennedy (1996). "Protection Against 
A Lethal Challenge With SV40-Transformed Cells By The Direct Injection Of DNA- 
Encoding SV40 Large Tumor Antigen." Cancer Res 56(5): 1 126-30. 

Bright, R. K., M. H. Shearer and R. C. Kennedy (1995). "Nucleic Acid Vaccination 
Against Virally Induced Tumors." Ann N Y Acad Sci 772: 241-51. 

Budker, V., G. Zhang, I. Danko, P. Williams and J. Wolff (1998). "The Efficient 
Expression Of Intravascularly Delivered DNA In Rat Muscle." Gene Ther 5(2): 272-6. 

Cantor, C. R., and Schimmel, P.R. (1980). Biophysical Chemistry Part III: The Behavior 
Of Biological Macromolecules . New York, W.H. Freeman And Company. 

Casares, S., K. Inaba, T. D. Brumeanu, R. M. Steinman and C. A. Bona (1997). "Antigen 
Presentation By Dendritic Cells After Immunization With DNA Encoding A Major 
Histocompatibility Complex Class Il-Restricted Viral Epitope." J Exp Med 186(9): 1481- 
6. 

Chemg, J. Y., SchuurmansNiewenbroek, Jiskoot, W., Talsma, N.J., Zuidam, W.E., 
Hennik, W.E., and Crommelin, D.J.A. (1999). "Effect Of DNA Topology On The 
Transfection Efficiency Of Poly((2-Dimethylamino)Ethyl Methacrylate)-Plasmid 
Complexes." J Cont Rel 60: 343-353. 

Chia, D., Dorsch, C.A., Levy, L., and Bamett, E.V. (1979). "The Metabolism Of Nucleic 
Acids In Mice." Immunology 36: 323-329. 

Chitrabamrung, S., Bannetu .I.S., Rubin, R.L., and Tan, E.M. (1981). "A Radial Diffusion 
Assay For Plasma And Serum Deoxyribonuclease I." Rheum Int 1(2): 49-53. 



159 



160 



Chused, T. M., Steinberg, A.D. and Talal, N. (1972). "The Clearance And Localization 
Of Nucleic Acids By New Zealand And Normal Mice." Clin Exper Imm 12: 465-476. 

Colote, S., Ferraz, C, and Liautrard, J.P. (1986). "Analysis And Purification Of Plasmid 
DNA By Reversed -Phase High-Perfomiance Liquid Chromatography." Analytical 
Biocheml54(l): 15-20. 

Corr, M., D. J. Lee, D. A. Carson and H. Tighe (1996). "Gene Vaccination With Naked 
Plasmid DNA: Mechanism Of CTL Priming." lEjcpJ^ed 184(4): 1555-60. 

Danko, I., P. Williams, H. Herweijer, G. Zhang, J. S. Latendresse, I. Bock and J. A. 
Wolff (1997). "High Expression Of Naked Plasmid DNA In Muscles Of Young 
Rodents." HumMoLGenet 6(9): 1435-43. 

Davies, B., and Morris, Tim (1993). "Physiological Parameters In Laboratory Animals 
And Humans." Pharm Res 10(7): 1093-1095. 

Dormelly, J. J., J. B. Ulmer and M. A. Liu (1995). "Protective Efficacy Of Intramuscular 
Immunization With Naked DNA." Ann N Y Acad Sci 772: 40-6. 

Dwyer, M. A., Huang, A.J., Pan, C.Q., and Lazarus, R.A. (1999). "Expression And 
Characterization Of A Dnase I-Fc Fusion Enzyme." J Biol Chem 274(14): 9738-9743. 

Etchart, N., R. Buckland, M. A. Liu, T. F. Wild and D. KaiserHan (1997). "Class I- 
Restricted CTL Induction By Mucosal Immunization With Naked DNA Encoding 
Measles Virus Haemagglulmin." .J Gen Virol 78(Pt 7): 1577-80. 

Fazio, V. M., S. Fazio, M. Rinaldi, M. V. Catani, S. Zotti, S. A. Ciafre, D. Seripa, G. 
Ricci and M. G. Farace (1994). "Accumulation Of Human Apolipoprotein-E In Rat 
Plasma After In Vivo Intramuscular Injection Of Naked DNA." Biochem Biophys Res 
Commun 200(1): 298-305. 

Gabizon, A., and Papahadjopoulos, D. (1988). "Liposome Formulations With Prolonged 
Circulation Time In Blood And Enhanced Uptake By Tumors." Proc Nat Acad Sci 85: 
6949-6953. 

Garrett, E. R. (1984). Integral Procedures In Pharmacokinetics And Their Application To 
The Parameter Determination Of Appearing Metabolites. Pharmacokinetics. A Modem 
View . L. Z. Levy. New York, Plenum Press: 253-280. 

Gibaldi, M., and Perrier, D. ( 1 982). Pharmacokinetics . New York, Marcel Dekker Inc. 

Gosse, C. H., Le Pecq, J.B., Defrance, P., and Padetti, C. (1965). "Initial Degradation Of 
Deoxyribonucleic Acid After Injection In Mammals." Cancer Res 25: 877. 



161 



Haensler, J., C. Verdelet, V. Sanchez, Y. Girerd-Chambaz, A. Bonnin, E. Trannoy, S. 
BCrishnan and P. Meulien ( 1 999). "Intradermal DNA Immunization By Using Jet- 
Injectors In Mice And Monkeys." Vaccine 17(7-8): 628-38. 

Hirose, S., and Ohaba, R. ( 1 993). Supercoiling Facilitates The Assembly Of Active 
Chromatin. Molecular Biology Of DNA Topoisomerases . T. Andoh, Dceda, H., And 
Oguro, M. Boca Raton, CRC Press: 87-93. 

Houk, B. E., Hochhaus, G., and Hughes, J. A. (1999). "Kinetic Modeling Of Plasmid 
DNA Degradation In Rat Plasma." PharmSci 1(4). 

Hwang, K. J. (1987). Liposome Pharmacokinetics. Liposomes From Biophysics To 
Therapeutics . M. J. Ostro. New York, Marcel Dekker: 109-156. 

Juliano, R. L. (1988). "Factors Affecting The Clearance Kinetics And Tissue Distribution 
Of Liposomes, Microspheres And Emulsions." Advan Drug Del Rev 2: 31-54. 

Katsumi, A., N. Emi, A. Abe, Y. Hasegawa, M. Ito and H. Saito (1994). "Humoral And 
Cellular Immunity To An Encoded Protein Induced By Direct DNA Injection." Hum 
Gene Ther Sn L): 1335-9. 

Katz, L., WilHams, P.H., Sato, S., Leavitt, R.W., and Donald, R.H. (1977). "Purification 
And Characterization Of Covalently Closed Replicative Intermiediates Of Colel DNA 
From Escherichia Coh." Biochemistry 16(8): 1677-1683. 

Kawabata, K., Y. Takakaura and M. Hashida (1995). "The Fate Of Plasmid DNA After 
Intravenous Injection In Mice: Involvement Of Scavenger Receptors In Its Hepatic 
Uptake." PharrnRes 12(6): 825-830. 

Lawson, C. M., W. S. Yeow, C. M. Lee and M. W. Beilharz (1997). "In Vivo Expression 
Of An Interferon-Alpha Gene By Intramuscular Injection Of Naked DNA." J Interferon 
Cytokine Res 17(5): 255-61 . 

Lew, D., S. E. Parker, T. Latimer, A. M. Abai, A. Kuwahara-Rundell, S. G. Doh, Z.-Y. 
Yang, D. Laface, S. H. Gromkowski, G. J. Nabel, M. Manthorpe and J. Norman (1995). 
"Cancer Gene Therapy Using Plasmid DNA: Pharmacokinetic Study Of DNA Following 
Injection In Mice." Hum Gene Ther 6: 553-564. 

Liu, F., Qi, H., Huang, L.. and Liu, D. (1 997). "Factors Controlling The Efficiency Of 
Cationic Lipid-Mediated Transfection In Vivo Via Intravenous Administration." Gene 
Ther 4: 517-523. 

Liu, F., Song, Y.K., and Liu. D. (1999). "Hydrodynamics-Based Transfection In Animals 
By Systemic Administration Of Plasmid DNA." Gene Ther 6: 1-9. 



162 



Lodish, H., Baltimore, D., Berk, A.,La\\ rcnce, S.Z., Matsudaira, P., and Darnell, J. 
(1995). Molecular Cell Biology . New York, Scientific American Books. 

Lopez-Macias, C, M. A. Lopez-Hernandez, C. R. Gonzalez, A. Isibasi and V. Ortiz- 
Navarrete (1995). "Induction Of Antibodies Against Salmonella Typhi Ompc Porin By 
Naked DNA Immunization." Ann N Y Acad Sci 772: 285-8. 

Mahato, R. I., Takakura, Y., and Hashida, M. (1997). "Nonviral Vectors For/« Vivo 
Gene Delivery: Physiochemical And Pharmacokinetic Considerations." Crit Rev Ther 
Drug Carrier Svs 14(2): 133-172. 

Mahato, R. I., K. Kawabala, T. Nomura, Y. Takakaura and M. Hashida (1995). 
"Physiochemical And Phamiacokinetid Characteristics Of Plasmid DNA/Cationic 
Liposome Complexes." .1 Pharm Sci 84(11 ): 1267-1271. 

Melzak, K. A., Sherwood, C.S., Turner, R.F.B., and Haynes, C.A. (1996). "Driving 
Forces Fro DNA Adsorption To Silica In Percholate Solutions." J Colloid Interface Sci 
181:635-644. 

Micromath (1995). Scientist For Experimental Data Fitting . Salt Lake City, Micromath. 

Murray, E. J., Ed. (1991). Meth Mol Bio . CUfton, NJ, The Humana Press Inc. 

Niven, R., Pearlman, R., Wedking, T., Mackeigan, J., Noker, P., Simpson-Herren, L., and 
Smith, J.G. (1998). "Biodistribution Of Radiolabeled Lipid-DNA Complexes And DNA 
In Mice." .1 Pharm Sci 87(11): 1292-1299. 

Noll, A., N. Bucheler, E. Bohn, R. Schirmbeck, J. Reimann and I. B. Autenrieth (1999). 
"DNA Immunization Confers Systemic, But Not Mucosal, Protection Against 
Enteroinvasive Bacteria." Eur J Immunol 29(3): 986-96. 

Osaka, G., K. Carey, A. Cuthbertson, P. Godowski, T. Patapoff, A. Ryan, T. Gadek and J. 
Mordenti (1996). "PhaiTnacokinetics, Tissue Distribution, And Expression Efficiency Of 
Plasmid [33P] DNA Following Intravenous Administration Of DNA/Cationic Lipid 
Complexes In Mice: Use Of A Novel Radionucleolide Approach." J Pharm Sci 85(6): 
612-618. 

Osorio, J. E., C. C. Tomlinson, R. S. Frank, E. J. Haanes, K. Rushlow, J. R. Haynes and 
D. T. Stinchcomb (1999). "Immunization Of Dogs And Cats With A DNA Vaccine 
Against Rabies Virus." Vaccine 17(9-10): 1 109-16. 

Piva, R., Lambertini, E., Penolazzi, L., Facciolo, M.C., Lodi, A., Aguiari, G., Nastruzzi, 
C, and Del Senno, L. (1998). "In Vitro Stability Of Polymerase Chain Reaction- 
Generated DNA Fragments In Serum And Cell Extracts." Biochem Pharmacol 56: 703- 
708. 



163 



Poly, F., Chenu, C, Simonet, P., Rouiller, J., and Monrozier, L.J. (1999). "Differences 
Between Linear Chromasomal And Supercoiled Plasmid DNA In Their Mechanisms And 
Extent Of Adsorption On Clay Minerals." Langmuir 15: Page Est: 5.8. 

Associated Press. (1999). Patient's Death Stops Gene Therapy Studies. New York Times. 
October. 

Ragno, S., M. J. Colston, D. B. Lowrie, V. R. Winrow, D. R. Blake and R. Tascon 
(1997). "Protection Of Rats From Adjuvant Arthritis By Immunization With Naked DNA 
Encoding For Mycobacterial Heat Shock Protein 65." Arthritis Rheum 40(2): 277-83. 

Ramsey, J. P., Clemet, R.L., Hulley, G.T., and Tarr, G.C. (1999). "Assessment Of In 
Vitro Expression Levels Of Plasmid DNA Conformers." PharmSci 1(4): S-4094. 

Rizzuto, G., M. Cappelletti, D. Maione, R. Savino, D. Lazzaro, P. Costa, I. Mathiesen, R. 
Cortese, G. Ciliberto, R. Laufer, N. La Monica and E. Fattori (1999). "Efficient And 
Regulated Erythropoietin Production By Naked DNA Injection And Muscle 
Electroporation." Proc Natl Acad Sci U S A 96(1 1): 6417-22. 

Sato, Y., Kiwada, H., and Kate, Y. (1986). "Effects Of Dose And Vesicle Size On The 
Pharmacokietics Of Liposomes." Chem Pharm Bull 34(10): 4244-4252. 

Song, Y. K., F. Liu, S. Chu and D. Liu (1997). "Characterization Of Cationic Liposome - 
Mediated Gene Transfer In Vivo By Intravenous Administration." Hum Gene Ther 8: 
1585-1594. 

Song, Y. K., F. Liu and D. Liu (1998). "Enhanced Gene Expression In Mouse Lung By 
Prolonging The Retention Time Of Intravenously Injected Plasmid DNA." Gene Ther 
5(11): 1531-7. 

Soubrane, C, R. Mouawad. O. Rixe, V. Calvez, A. Ghoumari, O. Verola, M. Weil and D. 
Khayat (1996). "Direct Gene Transfer Of A Plasmid Carrying The Herpes Simplex 
Virus- Thymidine Kinase Gene (HSV-TK) In Transplanted Murine Melanoma: In Vivo 
Study." Eur J Cancer 32A(4): 691-5. 

Takakura, Y. and Hashida, M. (1996). "Macromolecular Carrier Systems For Tageted 
Drug Delivery: Pharmacokinetic Considerations On Biodistribution." Pharm Res 13(6): 
820-831. 

Takeshita, S., Y. Tsurumi, T. Couffmahl, T. Asahara, C. Bauters, J. Symes, N. Ferrara 
and J. M. Isner (1996). "Gene Transfer Of Naked DNA Encoding For Three Isoforms Of 
Vascular Endothelial Growth Factor Stimulates Collateral Development In Vivo." Lab 
Invesl 75(4): 487-501. 



164 



Thierry, A. R., Lunardi-Iskandar, Y., Bryant, Joseph L., Rabinovich, Peter, Gallo, Robert 
C, And Mahan, and Lawrence C. (1995). "Systemic Gene Therapy: Biodistribution And 
Long-Term Expression Of A Transgene In Mice." Proc Natl Acad Sci 92: 9742-9746. 

Thierry, A. R., R. Rabinovich, B. Peng, L. C. Mahan, J. L. Bryant and R. C. Gallo 
(1997). "Characterization Of Liposome-Mediated Gene Delivery: Expression , Stability, 
And Pharmacokinetics Of Plasmid DNA." Gene Ther 4: 226-237. 

Tsumita, T. and Iwanaga, M. (1963). "Fate Of Injected Deoxyribonucleic Acid In Mice." 
Nature 198: 1088-1089. 

Ulmer, J. B., J. J. Donnelly. R. R. Deck, C. M. Dewitt and M. A. Liu (1995). 
"Immunization Against \iral Proteins With Naked DNA." Ann N Y Acad Sci 772: 117- 
25. 

Valere, T. (1999). "Gene Transfer And Therapy Clinical Trials-Update. Part 1: Countries, 
Diseases, Clinical Trial Phases, Routes Of Administration." J Gene Med 1(1): 71-73. 

Wang, C, C. Chao, P. Madeddu, L. Chao and J. Chao (1998). "Central Delivery Of 
Human Tissue Kallikrein Gene Reduces Blood Pressure In Hypertensive Rats." Biochem 
Biophys Res Commun 244(2): 449-54. 

Wang, C, L. Chao and J. Chao (1995). "Direct Gene Delivery Of Human Tissue 
Kallikrein Reduces Blood Pressure In Spontaneously Hypertensive Rats." J Clin Invest 
95(4): 1710-6. 

WTialey, K., and Webb, .!. (1972). "Clearance Of Deoxyribonucleic Acid From Blood In 
New Zealand Mice." Reviews De European Etudes Clinical Et Biologic 17: 788-791. 

Witzenbichler, B., T. Asahara, T. Murohara, M. Silver, I. Spyridopoulos, M. Magner, N. 
Principe, M. Kearney, J. S. Hu and J. M. Isner (1998). "Vascular Endothelial Growth 
Factor-C (VEGF-C/VEGF-2) Promotes Angiogenesis In The Setting Of Tissue 
Ischemia." Am J Pathol 153(2): 381-94. 

Wolff, J. A., P. Williams. G. Acsadi, S. Jiao, A. Jani and W. Chong (1991). "Conditions 
Affecting Direct Gene Transfer Into Rodent Muscle In Vivo." Biotechniques 11(4): 474- 
85. 

Yang, J. P. and L. Huang (1996). "Direct Gene Transfer To Mouse Melanoma By 
Intratumor Injection Of Free DNA." Gene Ther 3(6): 542-8. 

Yarmola, E. G., Zarudnaya, M.I., and Lazurkin, Y. S. (1985). "Osmotic Pressure Of 
DNA Solutions And Effecti\ e Diameter Of The Double Helix." J Of Biomol Struct Dyn 
2(5): 981-993. 



165 



Yoshida, M., Mahato, R.I., Kawabata, K., Takakura, Y., and Hashida, M. (1996). 
"Disposition Characteristics Of Plasmid DNA In The Single-Pass Rat Liver Perfusion 
System." PharmRes 13(4): 599-603. 

Zhang, G., V. Budker and J. A. Wolff (1999). "High Levels Of Foreign Gene Expression 
In Hepatocytes After Tail Vein Injections Of Naked Plasmid DNA." Hum Gene Ther 
10(10): 1735-7. 

Zhang, G., D. Vargo, V. Budker, N. Araistrong, S. Knechtle and J. A. Wolff (1997). 
"Expression Of Naked Plasmid DNA Injected Into The Afferent And Efferent Vessels Of 
Rodent And Dog Livers." Hum Gene Ther 8(15): 1763-72. 



BIOGRAPHICAL SKETCH 

Brett Edward Houk was bom on April 1 7th 1969 in Parma Ohio, and spent most 
of his childhood in Strongsville, Ohio. Brett attended the Ohio State University where he 
obtained Bachelor of Science degrees in both allied medicine and economics. After 
graduating from undergraduate studies, Brett moved to West Palm Beach, Florida and 
worked in Lantana Public Health Clinic in clinical pharmacokinetics. It was here that 
Brett decided to pursue a career as a pharmaceutical scientist. Brett was admitted to the 
University of Florida College of Pharmacy as a graduate student in August of 1996. 

In his spare time Brett enjoys reading, swimming, working out, and playing sports 
of all varieties. 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 





ie^^^A. Hughes, Chair 
associate Professor of Pharmaceutics 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 

<$>. /^ Li ^iC-c^-, 

Guenther Hochhaus, Cochair 
Associate Professor of Pharmaceutics 

I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 



U. l>^iQj^ 



Hartmut Derendorf 
Professor of Pharmaceutics 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 



P. 





Gayle\A. Brazeau 

Associate Professor of Pharmaceutics 



I certify that I have read this study and that in my opinion it conforms to 
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, 
as a dissertation for the degree of Doctor of Philosophy. 



ejL^rv^^-^ 



Edwin M. Meyer 
Associate Professor of Physiology and 
Pharmacology 



This dissertation was submitted to the Graduate Faculty of the College of 
Pharmacy and to the Graduate School and was accepted as partial fulfilhnenii 
requirements for the degree of Doctor of Philosophy. 

May 2000 

Pharmacy 




Dean, Graduate School 



UNIVERSITY OF FLORIDA 



3 1262 08555 2742